 Your new post is loading...
"Après l’ouverture, le 18 janvier dernier, des inscriptions préliminaires à la deuxième session du concours national de recrutement de 310 étudiants magistrats, appelés à exercer au sein des tribunaux et des cours, le ministère de la Justice a annoncé lundi, le début des inscriptions à la troisième session du concours national de recrutement des traducteurs-interprètes officiels. 80 postes vacants sont à pourvoir au profit des candidats titulaires d’un diplôme de licence ou de master en traduction délivré par un institut de traduction ou d’un diplôme étranger reconnu équivalent. Dans son communiqué rendu public, le ministère de la Justice informe les potentiels candidats désireux de faire carrière dans le domaine judiciaire que les inscriptions préalables sur la plateforme numérique débuteront à partir du 27 février et se poursuivront jusqu’au 5 mars 2025. Les inscriptions préalables s’effectueront par le lien suivant : http:/concours.mjustice.dz. Quant au dépôt des dossiers auprès des conseils judiciaires, l’opération s’étalera du 6 mars au 13 mars 2025. Les candidats dont les dossiers auront été validés sont avisés par ailleurs que les épreuves écrites du concours se dérouleront le 12 avril au niveau de l’institut de traduction de Bouzaréah-université d’Alger 2. Les conditions de participation au concours requises par le ministère de la Justice exigent la nationalité algérienne, le diplôme, une expérience d’au moins de 5 ans dans la profession de traducteur, dans un service de traduction au sein d’une juridiction, d’une administration, d’une institution, d’un établissement public ou privé, d’une organisation, d’un office public de traduction officielle ou d’un office étranger de traduction ; et enfin que le candidat soit âgé d’au moins 25 ans et jouit de ses droits civiques. En ce qui concerne le dossier administratif à fournir, il comprend un formulaire d’inscription rempli sur la plateforme électronique ; un extrait de l’acte de naissance ; une copie de la carte d’identité biométrique ; deux (2) photos d’identité ; une copie du diplôme universitaire requis ; une attestation de travail justifiant d’une ancienneté minimale de cinq (5) ans dans un service de traduction ; une attestation délivrée par l’administration ayant l’autorité de nomination, autorisant le candidat fonctionnaire à participer au concours et s’engageant à accepter sa démission en cas d’admission finale ; et une copie du mandat de versement des droits de participation au concours d’une valeur de 5000 DA. Le ministère de la Justice informe les candidats intéressés par ce concours que les inscriptions préalables sur la plateforme numérique s’effectueront par le biais du lien suivant : https:/concours.mjustice.dz ; pour de plus amples informations, les candidats sont invités à consulter le site : concours.mjustice.dz.
M.M"
https://lestrepublicain.com/2025/02/21/concours-pour-80-postes-de-traducteurs-interpretes-le-ministere-de-la-justice-recrute/
#metaglossia_mundus
In Bangladesh, visually challenged people have little access to the rich world of literature, an exclusion that limits their ability to experience the joy of storytelling.
"...We eagerly consume foreign literature in English or their Bangla translations, yet we show little curiosity about the literature written in our country's Indigenous languages. Should we not be equally invested in exploring our own linguistic heritage? The role of institutions is crucial here. The cultural affairs ministry, for instance, should push for a greater representation of Indigenous literature. Bangla Academy and the International Mother Language Institute should take the lead in publishing works in Indigenous languages and translating them into Bangla..."
Laila Khondkar
https://www.thedailystar.net/opinion/views/news/we-need-literature-all-people-all-languages-3830346
"New Delhi: Country's first-of-its-kind dictionary, which aims to bring uniformity in sign languages used by hearing and speech impaired people across the nation, was released today.
The Indian Sign Language (ISL) dictionary, which has been developed by the Indian Sign Language Research and Training Centre (ISLRTC), comprises 3,000 Hindi and English words and their corresponding graphic representation of the signs which are used in daily life.
The dictionary is available both in print and video format.
"The basic aim of developing the ISL Dictionary is to remove communication barriers between the deaf and hearing impaired communities as it is focused on providing more information in Indian sign language," Union Social Justice and Empowerment Minister Thaawarchand Gehlot said at the release of the dictionary.
He said that soon many new words will be added in the dictionary and hoped that more and more people will use this dictionary and get the benefit of learning Indian Sign language.
The dictionary contains graphic representations of popular signs used by the hearing impaired and also includes regional variations.
"It will also enable government officials, teachers, professionals, community leaders and the public at large to learn and use sign language," a senior ministry official said.
As per the census 2011, there are 50,71,007 deaf and 19,98,535 people with speech disability in India.
Around 300 dictionaries were distributed today. The DVD of the dictionary will also be uploaded on the ministry's website and people can procure it from ISLRTC too.
ISLRTC has been working on this ambitious project for the last one-and-half years.
The words in dictionary are divided into five five categories-- Everyday, Legal, Academic, Medical, and Technical terms, , the statement said.
Everyday terms category includes words that are used in everyday communication. Legal category includes videos for 237 legal terms and explains legal words like “affidavitâ€, “acquittalâ€, etc., that are used in various legal situations.
Academic terms help deaf children understand complex academic concepts and contains explanations for terms like "nervous system", "rotation" and "revolution", etc. The words are from various subjects like Physics, Geography, Biology, Maths, among others. This category contains 229 videos for 212 terms.
Medical category includes 200 sign videos for 200 medical terms to help the deaf community understand the medical terminologies used in hospitals and in medical situations.
Technical category has 206 videos of signs and explanations for 204 technical terms that are used in vocational training or in computer courses. "
https://www.dinamalar.com/news/kalvimalar-news-en/india39s-first-sign-language-dictionary-released/27856
#metaglossia_mundus
"International Mother Language Day marked in Ethiopia with call for multilingualism
Source: Xinhua| 2025-02-21 00:47:15|
Guests attend a celebration observing International Mother Language Day at the United Nations Conference Center in Addis Ababa, Ethiopia, Feb. 20, 2025. (Xinhua/Habtamu Worku)
ADDIS ABABA, Feb. 20 (Xinhua) -- Experts and policymakers in Ethiopia have called for redoubling concerted efforts in promoting linguistic diversity and multilingualism during a celebration observing International Mother Language Day.
On Thursday, a special event was held at the United Nations Conference Center in Addis Ababa, Ethiopia's capital, under the theme "Languages Matter: Silver Jubilee Celebration of International Mother Language Day."
Addressing the event, Aboubakri Diaw, chief of staff at the UN Economic Commission for Africa (UNECA), said the celebration "provides a unique opportunity to take stock of the progress made in promoting multilingual education, while recognizing the power of languages to bring individuals, people, and nations together."
Data from the UN Educational, Scientific and Cultural Organization (UNESCO) showed that around 40 percent of the estimated 8,324 languages worldwide are at risk of disappearing.
Abdoulaye Salifou, education program specialist at the UNESCO Liaison Office with the African Union (AU) and UNECA, highlighted the importance of promoting linguistic and cultural diversity as well as the crucial role of mother language in education, communication, culture, and science.
He Youlin, deputy head of the Chinese Mission to the AU, emphasized that the international community has an urgent responsibility to preserve and protect "precious" language resources.
"Language is a tool for human beings to exchange thoughts, a carrier of inheriting civilization, and a bridge for enhancing understanding, which emphasizes the importance of language. The celebration of the International Mother Language Day reminds us that protecting language diversity is protecting the common cultural heritage of mankind," he said.
He noted that China, home to 56 ethnic groups, launched the Chinese language resource protection project in 2015, which featured numerous surveys and efforts for the preservation, display, development, and utilization of linguistic resources.
International Mother Language Day, initially proclaimed by UNESCO and later adopted by the UN General Assembly, is observed annually on Feb. 21. ■"
https://english.news.cn/africa/20250221/cd3e06abdc8d4730b731e4052fde002b/c.html
#metaglossia_mundus
UCT PhD candidate Naledi Maponopono has been appointed as a member of Pan South African Language Board.
"To recognise her unending commitment to advancing indigenous African languages and fast-tracking multilingualism in South Africa, the minister of Sport, Arts and Culture, Gayton McKenzie, appointed Naledi Maponopono to the board of the Pan South African Language Board (PanSALB).
PanSALB is committed to promoting multilingualism in South Africa and developing and preserving the country’s 12 official languages. And appointing Maponopono – a PhD candidate in the University of Cape Town’s (UCT) African Languages and Literatures section in the School of Languages and Literatures – seems only fitting. She has dedicated her academic career to studying African languages and literatures.
Naledi Maponopono. Photo Supplied.
“PanSALB is a powerful instrument in our democracy, designed to safeguard our languages and I am honoured to contribute,” she said. “This appointment is a testament to my lifelong commitment to advancing indigenous African languages. From the time I started my academic career it was my goal to serve this board eventually.”
A powerful tool for unity
According to McKenzie, Maponopono will be required to provide oversight and strategic direction to the board – ensuring that it fulfils its constitutional mandate of promoting and developing previously marginalised official languages, Sign Language, as well as Khoi, Nama and San languages.
“She is expected to serve impartially and independently, and exercise, carry out and perform her powers, duties and functions in good faith and without fear, favour, bias or prejudice, subject only to the Constitution and the Act,” McKenzie stated in the letter of appointment.
“In a country like South Africa, with a history of radical injustice, language can be an instrument to bring people together.”
As Maponopono’s five-year term as a board member gets under way, she said she has always dreamed of playing a central role in shaping the linguistic landscape of the country, both through research and policy development. This appointment provides her with an opportunity to do so.
And once her term ends, she said, her goal is to be remembered as a board member who used language to unite South Africans under a common purpose: to make the country a better place.
“As the Sapir–Whorf hypothesis suggests, ‘Language shapes our world view’. Each time we learn a new language; we step into someone else’s world and experience their perspective. In a country like South Africa, with a history of radical injustice, language can be an instrument to bring people together,” she said."
https://www.news.uct.ac.za/article/-2025-02-20-language-can-be-an-instrument-to-bring-people-together
"After K-pop, K-movies, and K-dramas, K-LITERATURE is the latest form of K-culture gaining global attention. In the five-year period from 2018 to 2022, 1.85 million copies of Korean literature were sold overseas. As the genres preferred by global readers diversify, Korean literature introduced abroad has been expanding to include graphic novels, essays, and more. For today’s ‘Hallyu People’ we connect live to Sydney, Australia where Mark Oh, a young translator, introduces Korean FANTASY literature to international readers. Hello Mark!
1) Mark, please introduce yourself to our viewers and what you currently do in Australia.
2) You are translating into English the fantasy novel series “Pangaea” which has been written by a Korean author. Please tell us how this started.
3) It must be very special to translate your mother’s works. How do you feel about it and also how does your mother, author Ha Jeeyoon feel about it?
4) Is there anything in particular you pay attention to when translating the Pangaea series in terms of the uniqueness of fantasy fiction and also considering that your mother wrote them?
5) Do you have any special experience to share that helped you translate “The Pangaea” series?
6) How much exposure is Korean literature receiving in Australia?
7) The Pangaea novel series is also expected to be adapted into other types of content I heard?
8) What are your future plans or aspirations as a translator?...
#KLITERATURE #Korean_literature #Korean_fantasy_stories #Pangaea #MarkOh #한국판타지소설 #문학번역가 #Arirang_News #아리랑뉴스
📣 Facebook : / arirangtvnews
📣 Twitter : / arirangtvnews
📣 Homepage : https://arirang.com/
2025-02-20, 14:00 (KST)"
>>>>
https://www.google.com/url?rct=j&sa=t&url=https://www.youtube.com/watch%3Fv%3DMdvTQPt4_WQ&ct=ga&cd=CAEYBioUMTIzNzgyMDM4OTMyMDg2NTMzMTMyGjA5M2M4YjQ0YTI5OWZlNGE6Y29tOmVuOlVT&usg=AOvVaw1y-p20DrTTr35haI9A2WZu
#metaglossia_mundus: A good example for users of other 7000+ mother languages across the world...
"AUSIT NAT Presents: Simultaneous Interpreting in Conference-Focused Settings
Webinar Via Zoom
Wednesday 12 March 2025
7.00 pm - 9.00 pm AEDT
Registrations close 10 March 2025
Do you already use simultaneous interpreting in your work, or are you curious to find out more about this mode of interpreting?
Simultaneous interpreting (SI) plays a pivotal role in ensuring seamless communication and engagement at international conferences. As technology evolves, conference interpreters must adapt to new platforms, tools, and expectations.
In this session, we will cover the following topics:
An overview of simultaneous interpreting (SI) in the contemporary era;
The technological evolution and its impact upon SI;
The practice of SI in Australia: settings, stakeholders and pathways
How to maintain/upgrade one’s SI competence for conferences
The use of AI in SI skills development via deliberate practice
Q&A
Have a question about SI? Click here to submit it!
Stay tuned for our second episode where we will explore SI in community settings.
Our Presenter - Shiyi Ye
Shiyi Ye holds a Master of Conference Interpreting from Macquarie University in Sydney, where he teaches interpreting and translation part-time. He completed his training in interpreter education at the University of Geneva in Switzerland. In addition to training future colleagues, he continues to work as a freelance interpreter and translator, serving various events and clients in Australia and overseas.
A full member of AUSIT since 2010, he has served in various roles on the NSW Branch Committee, including Branch Secretary and Branch Delegate. He is also an active member of both the International Association of Conference Interpreters and the Translators Association of China.
Joining the Webinar
The Zoom links for the webinar will be sent out 1 day prior to the event. Please ensure you have downloaded the latest version of Zoom in preparation.
PD Logbook
Introductory PD.
Please consult the NAATI Recertification Catalogue to allocate the relevant PD Points. Some activities can be included under different categories, so please choose the most appropriate category for your recertification purposes.
Certificates and Recordings
Webinar recordings and Certificates for PD purposes may take approximately two (2) weeks to be released following the event.
This is a live webinar event and will be recorded. The recording
will be shared with registrants.
Wednesday 12 March 2025
7:00 pm - 9:00 pm AEDT
Click here to view time zone converter
Zoom online Webinar
AUSIT Members - AUD $30
AUSIT Students - AUD $15
Members of ASLITA, NZSTI and associations that are
FIT-ITF members - AUD $30
Non-Members - AUD $90
Register>>>>
https://office.ausit.org/event-6072241
#metaglossia_mundus
Turkey’s translators are training the AI tools that will replace them
Human fallout may include being replaced by LLMs, diminished skills, and fewer career options for all but the elite scholars.
20 FEBRUARY 2025 • ISTANBUL, TURKEY
- As AI improves, translation jobs in Turkey have become scarce.
- Many young translators are editing AI-generated content or training AI to master Turkish.
- Their training helps LLMs powering Meta AI and OpenAI’s ChatGPT, among others.
Turkey’s sophisticated translators are moonlighting as trainers of artificial intelligence models, even as their profession shrinks with the rise of machine translations. As the models improve, these training jobs, too, may disappear. “Translation has become a job for a limited number of translators who are really good at it. Machines translate the rest,” Mehmet Şahin, head of the translation department at Boğaziçi University in Istanbul, told Rest of World. Translators will become less skilled in the future as AI eats up entry-level work, he said. Mehmet Şahin, head of the translation department at Boğaziçi University in İstanbul.As a teenager, Pelin Türkmen dreamed of becoming an interpreter, translating English into Turkish, and vice versa, in real time. She imagined jet-setting around the world with diplomats and scholars, and participating in history-making events. Her tasks one recent January morning didn’t figure in her dreams. The 28-year-old translator’s computer displayed a dashboard for AI training provided by Outlier, a San Francisco-based company that hires contractors to train large language models. Outlier’s clients include OpenAI, Microsoft, and Meta, among others. The dashboard displayed a prompt for training an LLM powering either ChatGPT, Gemini, or Preplexity AI. The AI chatbot provided three menus. Türkmen rated them on accuracy, ethics, and relevance. She checked for grammar, fluency, tone, and structure. She looked for language that would reveal an AI author — phrases like “As an AI assistant …” or “Certainly!” She explained her reasoning so that the machine could learn. Türkmen has earned several thousand dollars over the past nine months training AI. In the tech industry, this kind of work is known as reinforcement learning from human feedback, and it helps LLMs respond with intuition and context. “The answers they [LLMs] provide to Turkish prompts often sound machine-like,” Zeynep Kırıcı, another translator who contracts for Outlier, told Rest of World. “We’re trying to fine-tune the mechanical tone of AI. We’re trying to humanize it.” The job offers Türkmen a reliable income at a time when her career path looks like it’s disappearing, she said. Translators and scholars in Turkey told Rest of World the nature of translation work has changed. Before AI, young translators worked in translation bureaus tackling everything from administrative documents to trade reports and literary classics. Today, most entry-level positions involve editing AI-generated content as a machine translation “post editor,” they said. Others involve training AI. Şahin discusses the history of machine translation with his students.The shift is profound in Turkey, where translators have helped shape the nation’s culture and politics. After Turkey became a European-style republic a century ago, a government-sponsored translation bureau produced thousands of Western classics between 1940 and 1966, helping spread European ideas in society. Translators are highly esteemed and influential, and there were about 10,000 translators and 2,900 interpreters in 2015, according to the most recent data available from the Turkish government. Book publishing, too, is transforming. Turkish publisher Dedalus announced in 2023 that it had machine-translated nine books. In 2022, Agora Books, helmed by translator Osman Akınhay, released a Turkish edition of Jean-Dominique Brierre’s Milan Kundera, une vie d’écrivain, a biography of the Czech-French novelist Milan Kundera. Akınhay, who does not know French, used Google Translate to help him in the translation, to much criticism from the industry. “Nobody criticizes doctors who use Google Translate when talking to their patients,” Akınhay told Rest of World. “Nobody complains while using self-checkout desks in supermarkets.” The trends are global. In November, the Netherlands’ largest publisher announced it would use AI to translate certain books, with humans only involved in light edits. In the U.K., over a third of 12,500 translators said they have lost work to machines. “Frankly, we are at a stage where the profession of translator has lost its immunity,” Akınhay said. Outlier and its parent company, Scale AI, did not respond to a request for comment about their AI training operations in Turkey. San Francisco-based Outlier hires contractors to train large language models, for which they are paid weekly.The new roles require much less skill and effort than translation, Türkmen said. For instance, she spent a year on her master’s thesis studying Samuel Beckett’s self-translation of his play Endgame from French to English. More recently, for her Ph.D. in translation studies, she studied for more than two years about the anti-feminist discourse in the Turkish translation of French author Pierre Loti’s 1906 novel, Les Désenchantées. In contrast, working on an AI prompt takes about 20 minutes. For every hour Türkmen works, she gets $20. That’s as much as a translator earns after spending roughly four hours converting two pages of densely written text into Turkish, according to Çevbir, the association of Turkish translators. Outlier pays weekly, which is crucial as the Turkish lira fluctuates wildly. For book translations, workers are compensated only after publication — it can take months during which the payment can lose value. Globally, Outlier has a mixed reputation and Scale AI faces a class action lawsuit in the U.S. alleging non-payment by contractors. The company has responded to the suit, saying it complies with all laws and regulations and has safeguards to protect its workers. Erdem Hürer, 32, a graduate student in Boğaziçi University’s translation department and an Outlier contractor, told Rest of World he’d prefer to make his money entirely through AI training, while doing translations as a labor of love. Erdem Hürer is a graduate student at Boğaziçi University and works as an Outlier contractor.“[Translation is] an arduous job with so many details and labor involved that it drains away your time,” he said. The contractors said they feared that AI training jobs, which are plentiful now, will disappear once LLMs improve. Get The Global, our (free) newsletterSign up for our weekly newsletter and we’ll send you our latest stories, dispatches from our staff, what we’re reading, and more. A world of tech, right in your inbox. But that future is a few years away for Turkish-language LLMs, which don’t have much training data and rely on humans for fine-tuning, said Şahin. Yiğit Bener, former director of the Conference Interpreters Association of Turkey, believes AI will catch up and displace human interpreters in all except the most sensitive diplomatic meetings. Bener, 66, recounted live-translating a speech by the late Cuban President Fidel Castro when he visited Istanbul in 1996 for a United Nations conference. Castro spoke in a roaring tone, Bener recalled, then wondered whether AI could handle such a challenging meeting. He answered his own question. “In a year or two, as AI will offer flawless simultaneous translations, mediocre translators will be eliminated from the market,” he said. “And that is a good thing.”
Kaya Genç is a labor x tech reporter based in Istanbul, Turkey.
Whatever role is best for you—data scientist, data engineer, or technology manager—Northwestern University's MS in Data Science program will help you to prepare for the jobs of today and the jobs of the future.
Hal Varian, the chief economist at Google, is famous for saying in 2010 that “the sexy job in the next 10 years will be statisticians.” Building on this idea, Thomas Davenport and D.J. Patil wrote "Data Scientist: The Sexiest Job of the 21st Century."
Should you seek training in data science because the job is “sexy”? Probably not. You want to know that the job is in demand. Data scientists are in demand and so are data engineers.
Data scientists continue to be one of the fastest-growing occupations in the United States. According to the Bureau of Labor Statistics, the data scientist role boasts the 4th highest projected growth rate at 36% over the next decade (2023-2033), with an annual median pay of $108,020 in 2023.
There is also a high demand for data engineers. Indeed has included data engineers in their 2024 List of 10 Best Jobs in the US, with an average annual salary of $130,135.
The rise of generative AI and large language models has fundamentally transformed the technology landscape. Data scientist and data engineer roles are more critical than ever, particularly as organizations increasingly invest in these technologies. A 2024 Forrester survey reveals that 67% of AI decision-makers plan to increase generative AI investments in the coming year. This surge underscores the critical need for highly adaptable data scientists and data engineers with specialized skills in AI.
The Work of Data Science and Data Engineering
To remain competitive in today’s world, companies need to be informed by data and to use data in their day-to-day operations. That does not happen unless there are data engineers, technical professionals (software engineers, database administrators, cloud architects, and the like). These are people who can translate research results and data science models into systems that work.
The role of the data scientist is evolving with the emergence of capable generative AI and large language models. In addition to building new models from scratch, data scientists are integrating prebuilt models into the data pipeline. Many of these prebuilt models come from large technology firms. Building integrated, intelligent systems is the work of data engineering.
Data engineers have a key role to play at the beginning of every data science project. Data scientists depend on data engineers to gather and prepare data for analysis. Without data, there are no analyses, and no models to build and test.
Some analytics and modeling projects end with a written report to management or a display of results in a dashboard or presentation. Research findings guide management decisions.
Analysis and modeling projects need not end with a report. Many models are put into practice. They become the way a company conducts its business. Data engineers have key roles to play in building data science applications and implementing information systems.
Choosing the Right Job for You
Data scientists are like chameleons, changing their colors to match the business context. There are data scientists who could just as easily be called marketing researchers, financial analysts, or competitive intelligence professionals, depending on the work that needs to be done.
Analytics and modeling lie at the heart of data science, and many data scientists think of themselves as applied statisticians.
To serve the needs of today’s data-driven, data-intensive world, data scientists need to be multilingual, speaking the languages of information technology and business, as well as analytics and modeling.
Students interested in data science application development and systems implementation can specialize in data engineering and seek various information technology positions. Data engineers could just as easily be called software engineers, systems engineers, cloud architects, computer scientists, machine learning engineers, AI engineers, or development and operations (DevOpps) professionals. Job titles may vary from one company or one industry to the next, but the job opportunities are plentiful.
Reading job postings for data scientists and data engineers, you will find considerable overlap in the desired skillsets.
Data Scientist skills:
Programming
Cloud computing
Database management
Data visualization
Probability and statistics
Multivariate calculus and linear algebra
Machine learning & deep learning
Data Engineer skills:
Programming
Extraction Transformation and Loading (ETL)
System architecture
Database design and configuration
Interface and sensor configuration
Technology Management
Moving from the technology side to the management side of data science, presents additional opportunities. Organizations need people who can understand data science and data engineering, but also speak the language of business.
As with data science and data engineering, there are many jobs for technology managers. Someone needs to build and manage teams of data scientists and data engineers. Someone needs to make decisions about information infrastructure.
Communication skills are essential to technology management. There is a great need for people who can understand business problems as well as technical solutions to problems. There is a great need for people who can translate technical jargon into language that non-technologists can understand.
Whatever role is best for you—data scientist, data engineer, or technology manager—Northwestern University's MS in Data Science program will help you to prepare for the jobs of today and the jobs of the future. The program is offered in a convenient and engaging online format. Students learn part-time at their pace, allowing for learning and growth without career interruption.
https://www.kdnuggets.com/2025/02/nwu/data-scientist-data-engineer-or-technology-manager-which-job-is-right-for-you
"Last year, Elon Musk, CEO of Tesla, X, and SpaceX and a member of the Trump administration team, founded a startup called xAI to challenge OpenAI’s dominance of ChatGPT. While ChatGPT and other AI models are by far the most popular, Musk and his team released the latest version of their AI, Grok-3.
What is Grok-3?
Grok is the AI developed by xAI, and Grok-3 is the third and most sophisticated model yet. In a live demo on X, Musk said that AI is “an order of magnitude more capable than Grok 2 in a very short period of time” and that users can expect improvements to be made “literally every day.”
Musk and his team trained Grok-3 to have 10 times the computer power than Grok-2. That processing power will allow Grok-3 to outperform rivals like ChatGPT, DeepSeek, Gemini, and Copilot. However, it’s not just Musk’s team that believes the AI is superior to its rivals. The Artificial Analysis Quality Index puts Grok-3 above all the competitors mentioned above. Regarding computer power, Grok-3 is only second to OpenAI’s o3 program. O3 is a premium version of ChatGPT, which requires a subscription.
Additionally, Musk and his team at xAI plan to launch a new advanced search product, Deep Search. It’s supposed to be the “next-generation search engine” and will likely have to prove itself against Google Search. When speaking of this, Musk highlighted the unbelievable capability of Deep Search. Musk said, “Something that might take you half an hour or an hour of researching on the web or searching social media, you can just ask it to go do that and come back, and 10 minutes later, it’s done an hour’s worth of work to you… And maybe better than you could have done yourself.”
How Popular is this new AI?
Grok-3 certainly is powerful, but it isn’t perfect. Musk called the program a “maximally truth-seeking AI.” This claim, however, gets complicated when things get political. Musk has criticized Google Gemini and Chat GPT for being too “woke.” Musk and his team are marketing Grok-3 as an edgy and “able to answer spicy questions” that other AI models refuse to answer.
Like other AI models, many software users quickly noticed particular political leanings. Some suspect the software to reflect Musk’s right-leaning views. Musk claimed that the AI could help fight against the “woke mind virus.” Not everyone seems to be a fan of Grok-3 when it comes to politics. However, it is still true that the software has much more computing power than many of its competitors.
It’s hard to tell exactly how popular this new AI will be, but it certainly is shaking up the political, financial, and technology world. To stay updated on the latest financial news and keep up with Musk’s startup, be sure to tune into Due.com’s blog."
https://www.entrepreneur.com/finance/elon-musk-challenges-chatgpt-with-grok-3/487475
n#metaglossia_mundus
"L’association pour la promotion du français dans la mode veut mettre fin aux anglicismes
par Sharon Camara
20 févr. 2025
Barbiecore, royalecore, cowboycore, ces dernières années, les termes formés avec la terminaison anglaise « core » se multiplient. Une tendance que la Commission d’enrichissement de la langue française et l'Association pour la promotion du français dans la mode ne veulent plus voir.
« Gorpcore ? Dites plutôt : style randonneur ! », sur son compte Linkedin, l’association pour la promotion du français dans la mode invite les marques de mode à délaisser les anglicismes pour privilégier les termes français.
L’association base ses propos sur une publication au Journal officiel du 13 février d'une recommandation sur les équivalents français à donner aux termes formés avec « core ». Pour la Commission d’enrichissement de la langue française, ces termes sont utilisés « pour décrire un style vestimentaire et, par extension, un style de vie inspirés de la vision idéalisée d’un univers particulier », explique-t-elle.
« Cottagecore, royalcore, barbiecore, ou encore gorpcore, il est possible en français d’utiliser le mot style, et de parler ainsi de style inspiré de la campagne ou encore de style campagne, de style princier ou princesse, de style Barbie, et de style randonneur », affirme la commission.
L’association pour la promotion du français dans la mode invite les concernés (professionnels et passionnés de mode) à se rendre sur le site FranceTerme du ministère de la Culture qui a pour mission de promouvoir l’utilisation de mots en français et qui propose des équivalents en français de néologismes issus d’autres langues."
https://fashionunited.fr/actualite/culturelle/lassociation-pour-la-promotion-du-francais-dans-la-mode-veut-mettre-fin-aux-anglicismes/2025022037327
#metaglossia_mundus
Translation remains under-appreciated effect, says Telugu writers
"International Mother Language Day. Hyderabad: In a world where words bridge cultures and open minds, Telugu literature is caught in a quiet struggle. With an ever-growing appetite for translated works, the art of translation is still undervalued, leaving many powerful voices unheard. Telugu literature has witnessed a surge in translations both to and from the language, with notable names like V. Narayana Rao, Volga, Alladi Uma, M. Sridhar, and others leading the way. Also Read - AI to assist visually impaired students in Telangana Advertisement In recent years, translations from English and other Indian languages into Telugu have gained dedicated attention, particularly in the realm of fiction.
However, translation is not an easy job, especially when it is regarded as a part-time endeavour or hobby. "The job is still seen as a labour of love," says Purnima Tamireddy, a translator and publisher. "It’s difficult for publishers to compensate translators adequately, which doesn't align with the actual effort involved."
The challenges of translating books are manifold. One must consider not only the language but also the cultural contexts that both languages are grounded in. “You have to retain the local flavor without losing the essence of the original,” says Volga, a renowned Telugu author and translator. "When translating from English to Telugu, I choose authors I connect with emotionally, politically, and ideologically," she adds.
Telangana HC calls for compliance report on residential schools Remaining faithful to the original text is paramount, but Volga also notes that the greatest challenge remains finding the right words and phrases that capture the intent of the original author while resonating with the target audience. She also highlights a glaring gap in translations between Telugu and other languages. "If a hundred books are translated into Telugu from other languages, only about five are translated from Telugu to other languages," she mentions.
In recent times, sub-titling has emerged as a significant part of the translation business. “Although some attention is being given to it now, the quality of subtitles often remains poor,” says D.V. Subhashri, a professional freelance translator. "Time and space constraints make it difficult, and finding skilled sub-titlers is not easy." Languages are deeply entwined with culture, and translating them involves more than just words or phrases. "This is especially true with idioms or references to mythology. Indian readers may be familiar with figures like Sita and Rama, but these may require some explanation for international audiences," Subhashri adds. While certain genres have large followings, the tradition of "good-old storytelling" has largely taken a backseat, she notes.
Tamireddy further elaborates on the types of writings that attract readers: "When translated from Telugu to English, short stories and poetry are often overlooked. No one will read you unless you're a well-established author or have a significant social media presence," she says. "Novels tend to generate the most interest. Recently, anthologies have also gained popularity. An author’s single short story is more likely to make it into an anthology than a full collection by the same author," she adds. In addition to feminist and Dalit writings, science fiction and speculative fiction are also attracting growing interest.
https://www.deccanchronicle.com/southern-states/telangana/international-mother-language-day-is-on-february-21-1862188
"La la Faculté de traduction et d'interprétation (FTI) de l'Université de Genève a le plaisir d’annoncer que les inscriptions sont désormais ouvertes pour le programme Research Methods in Translation and Interpreting Studies (ReMeTIS).
Ce programme entièrement en ligne s'adresse aux personnes en doctorat et en pré-doctorat, aux chercheur·e·s dans des domaines connexes, ainsi qu'aux professionnel·le·s de la traduction et de l’interprétation souhaitant renforcer leurs compétences en recherche.
Il se compose désormais de quatre cours indépendants d'un mois (chacun nécessitant 50 à 60 heures d'apprentissage), permettant aux participant·e·s d'adapter leur parcours de formation à leurs besoins spécifiques.
Cours 1 : Exploration des études de traduction et d'interprétation - Questions de recherche et éthique (17 mars - 17 avril)
Cours 2 : Méthodes qualitatives - Conception et analyse de la recherche qualitative (15 mai - 15 juin)
Cours 3 : Méthodes quantitatives - Analyse statistique, conception d'enquêtes et interprétation des données (15 septembre - 15 octobre)
Cours 4 : Études de corpus - Compilation, analyse et interprétation de corpus pour explorer des modèles linguistiques et des phénomènes de traduction (17 novembre - 17 décembre)
Chaque cours suivi donne droit à 2 crédits ECTS, avec un certificat officiel de l'Université de Genève. La réussite des quatre cours donne droit à un total de 8 crédits ECTS et à un certificat de fin de programme.
Délais d'inscription :
Cours 1 : 24 février
Cours 2 : 24 avril
Cours 3 : 25 août
Cours 4 : 27 octobre
Pour plus de détails et d'informations sur l'inscription, visitez notre site web..."
https://www.unige.ch/fti/en/a-la-une/remetis-course-1/
With a passion for languages, Ray switched from economics to pursue a career in professional translation with help from MA Interpreting and Translating at Bath.
Bridging cultures through language: the art of interpretation
With a passion for languages, Ray switched from economics to pursue a career in professional translation with help from MA Interpreting and Translating at Bath.
Ray now works in the communications department of pharmaceutical company, Merck, Sharp & Dohme (MSD).
Growing up in China, Ray always had an affinity for languages. After graduating with an economics degree, he started working in the translation sector and knew he wanted to pursue this, leading him to study MA Interpreting and Translating (Chinese) at Bath.
He spoke to us about what led him to the course, the skills he gained, and how it unlocked doors for him in his career as a professional translator in China.
A life-long love of language
I remember my dad saying that language was the key to understanding the world. Even at an early age, that stuck with me and led me to become fascinated with the English language.
In my culture, it is common to study subjects like law or finance. So, at undergraduate level, I opted for economics at a top Beijing university; it was not something I found easy.
In my final year, by pure chance, I landed an internship at the Beijing office of The Wall Street Journal, where I translated English news stories, by brilliant reporters, into Chinese, for local and global Chinese readership.
I went into the internship with no prior experience in translation and no idea how well I could deliver. I still remember, on the last day of that first week, the department head called me to his office and told me that the editors found my work ‘unexpectedly good’. For me, after struggling for years academically, this was some long-awaited validation, and it made me realise for the first time that I might be able to do something I enjoyed doing for a living! The internship eventually turned into a full-time position at the Journal.
The next step for me was to find a course in interpretation and translation, and I knew I wanted to study abroad. Bath was a stand-out, with a reputation as one of the best universities for this subject. I had to sit a rigorous entry exam at the British Council in Beijing. I was nervous and after the exam, I drove around to try and stay calm whilst waiting for my results.
When I found out that I had passed I was so emotional and happy!
The work ethic and rigour of the course
Employers have a high regard for graduates from this course in Bath, as it has a reputation for delivering and training the best linguists, evident in the roles alumni have gone on to take.
With simultaneous interpreting, several linguistic processes take place at once, your mind must split into listening to what is being said, interpreting meaning and context, translating it into the correct language, and then speaking it, all at the point it happens.
There is no going back to correct it; accuracy and attention to detail is essential and so rewarding when you perfect this craft. It is challenging and so rewarding.
The class sizes are good for individual teaching and practice. There is plenty of time for academics to work with students, through labs, recordings, individual feedback, and a lot of practical work.
My cohort developed a very strong, self-driven atmosphere within the course; we booked sessions into our ‘free time’ to do further work, like reading the Economist cover to cover for example; we organised self-study groups and we would find cafés and kitchens and take turn to host and progress our work – it was very productive and collaborative.
‘I’ve now had a decade of experience in interpreting and translating and I still see a huge gap between what AI can do and what a top linguist can deliver. It is about empathy, about predicting, about contextualising, about understanding.’
—Ray ZhouMA Interpreting and Translating graduate (2018)
Feeling at home on campus and in the city
I was happy to be at a campus university; it is a helpful transition to living away from home.
I didn’t want to be in a big city, like London, it is too chaotic! The city and the campus at Bath are very compact. There are Chinese and other supermarkets too, so you can get your favourite condiments and ingredients.
The city itself is so beautiful and so easy to walk around. I have a playlist of songs I listened to when I was living there and when I’m missing Bath, I play them, and it takes me right back! I found it easy to get in and out of town, the bus service is very frequent and runs all day and night.
On a sunny day, everyone heads for the lake on campus and lies on the grass looking at the sky and the ducks, it is very relaxing and friendly. There is also a huge choice of groups and societies to join. The different interest groups set up little kiosks at a fair, answer questions and hand over information. I have such affection for, and feel so proud of, Bath and my time there. It has been the best decision of my life. I plan to bring my husband back to visit here one day, it is a beautiful destination.
The future is bright
My route into the profession was remarkably simple. I wanted to interpret, and I was prepared to study hard. I wanted to be great at my craft.
I worked at McKinsey in China as part of a big in-house translation team after they approached Bath directly for students. Employers know that Bath will produce students with exceptional qualities. My teachers put me forward for the interview and I was accepted!
I have now moved to work in the communications department of pharmaceutical company, Merck, Sharp & Dohme (MSD), where my job revolves around external and internal communications for the company leadership, with interpreting being an essential part of my work.
I’ve now had a decade of experience in interpreting and translating and I still see a huge gap between what artificial intelligence can do and what a top linguist can deliver. It is about empathy, about predicting, about contextualising, about understanding; you have to totally trust that the interpreter is going to empathise, and going to be able to reproduce your thoughts, feelings, emotions, and message accurately. Brilliant interpreters can sometimes speak the same word as the speaker in the same instant.
Tips to make the most of your time at Bath
Take advantage of how nurturing and supportive your time at university is, it is much harder in the real world with noises and distractions. Having the opportunity to listen and focus without those distractions is a gentle way into this profession, even though it can be tough.
This course is one where learning and improvement are done through constant constructive feedback and constructive criticism. It can be tough sometimes, but to perfect our craft, detailed feedback is essential - relish this opportunity to learn and grow.
Bath for me was outstanding. The welcome is polite, friendly, and solution-oriented, from academic support to mental health. There is some counselling offered to all students, maybe a few free sessions, and I thought why not use it if it is there, and it was really helpful.
The Wild Café in Bath was a favourite spot for me, I don’t think I’ve eaten better eggs benedict anywhere since!
One word of warning – it can get cold in the UK, so remember to pack some long underpants. That's important!...
Education experts have called on the government to integrate the teaching of vernacular languages in primary schools in efforts to promote local cultures and enhance comprehension among young learners.
The scholars said the tendency to favour Western languages like English in teaching and learning has been a hindrance to easy comprehension of some subjects, leading to poor performance.
Speaking during a book launch at Zetech University, Ruiru Campus, the experts said the languages should be taught in rural schools up to class six.
“We should revert to the old ways where learning vernacular languages had been integrated into the school curriculum. It enabled many learners to understand more about their cultures as well as helped in comprehension of even difficult subjects,” said Dr. Theophilus Kyando, a lecturer from the Technical University of Kenya.
Dr. Kyando argued that most students struggle in key subjects like mathematics and science, attributing these challenges to language barriers. He added that some children rarely understand their indigenous languages, which limits their creativity and comprehension of issues.
“This tendency to favour Western languages in learning has eroded our cultures. Even Western nations have preserved and even globalised their native languages, while African languages are often sidelined. This needs to stop,” said Dr. Kyando.
Prof. Oyier Nyaruath from the University of Nairobi echoed these sentiments, emphasising the need to revive a culture of reading. He cited Sweden as an example of a country that once deprioritized books in favour of digital media, only to later recognise the knowledge gap and reinstate book reading and writing.
Zetech University Vice-Chancellor, Prof. Njenga Munene, highlighted the significance of such forums in fostering intellectual growth. He noted that inviting experts from various institutions enriches academic discourse and sharpens students’ minds, making such engagements invaluable to any serious university.
The discussion took place during the launch of “I”, a book authored by Dr. Daniel Mwai that has details on human nature.
By Muoki Charles
Why mother languages matter
Ahead of International Mother Language Day, Newcastle-based Ulfet Sahin explains why embracing homegrown culture and tradition holds the key to happiness and confidence for thousands of young people Ulfet Sahin arrived on Tyneside eight years ago when her husband was awarded a fully funded opportunity to study for his PhD at Northumbria University. “I came from Turkey able to speak English and German — as well as my mother language — and we settled in quite quickly,” she explains. “My husband and I have visited many cities in the UK but we loved everything about Newcastle and knew that’s where we wanted to make our home. Once I was here I realised there was a need to help young Turkish people learn a new language, without losing ties to their native tongue and culture.” Ulfet is both a keen advocate of EAL — teaching English as an Additional Language — and an ardent supporter of retaining the centuries-old cultural ties that still bind communities and generations together, wherever they are in the world. Actively involved in International Mother Language Day (February 21), through her work with Success4All and Express Yourself: North East Festival of Languages, she understands better than most the value of language and cultural heritage. “I was a stammerer when I was a kid and it was really hard to talk my own language,” she adds. “But thanks to my great English teacher, I learned how to speak properly in my own language and in English. “Learning a second language is a bridge for kids between retaining key elements of their own culture and embracing a new culture. It’s an opportunity to express how you feel and understand how different languages — and dialects — can lead to different interpretations. “I’m a huge advocate for children arriving here learning English as soon as possible but never to the detriment of leaving their own language behind.” Ulfet, a tutor at educati A session from a previous Mother Language Day at the City Library in Newcastle on charity Success4All, works with the North East Festival of Languages team to promote Home Heritage and Community Languages (HHCL). This year’s festival workshops include Learning Arabic with Team Games and Yoruba language and culture events. Schools can also access a downloadable activity pack for Celebrating Indigenous African Languages. Ulfet has translated this year’s Festival Song — What A Wonderful World — into Turkish so that children can take part in a regional multilingual performance of the Louis Armstrong classic. And on February 24, she will shine a light on International Mother Language Day by reading stories in Turkish to children invited to attend Newcastle City Library’s free Love Your Mother Language Day. “To be honest, I feel much safer and more confident when I speak Turkish because it's my culture and it's my identity,” adds Ulfet. “And when I speak Turkish, I feel as if I’m at home. Last year, for example, we celebrated our Republic Day with many families now based in the North East. “People say the Turkish language sounds romantic and I love that. I know when and where to stress certain words that just don’t translate in the same way. That in itself sparks conversations around language that can lead to a greater understanding and appreciation of diverse cultures.” Ulfet has translated this year’s Festival Song — What A Wonderful World — into Turkish It was Bangladesh that first made moves to celebrate International Mother Language Day and an increasingly popular initiative has been observed worldwide since 2000. Backed by UNESCO, who believe in the importance of cultural and linguistic diversity in order to drive sustainable societies, the annual event is seen as a showcase for preserving the differences in cultures and languages that foster tolerance and respect for others. The Bangladeshi language, culture and traditions are celebrated at two in-person Express Yourself events in Newcastle and Sunderland in March — run by community associations and reinforcing the key message of International Mother Language Day. Nurturing HHCL offers significant social, economic, and cultural benefits to children and young people, investing in their personal development, future success, and the enrichment of society. And research shows that by supporting HHCL, schools not only empower individual students but also create a richer, more connected and globally minded school community. “In Turkey, I was a teacher,” adds Ulfet. “I taught English literature for 15 years but unfortunately, because of the education system over there, I couldn't work as a full-time teacher there. “Now I'm trying my best to help children coming over here to have the best opportunities possible. I come from a poor family and my mother was a single parent. That was tough but what I had in my favour was a good education. “Learning different languages opens so many doors and I urge my students not to close their minds to the opportunities that arise if you’re multilingual. “Thanks to learning English I was able to get a great job over here and they can do the same. But I also say ‘please keep your own language and your own culture’. One day that will be useful too.” Unfortunately for Ulfet, that reality is rarely reinforced in schools. “The problem is that most schools still don’t support home languages,” she explains. “They might offer French, German and Spanish but nothing else. Read more: Newcastle Theatre Royal’s Open Door initiative to nurture North East talent “I have a student who moved from Turkey just four months ago and he’s in Year 11. He's trying his best to learn English but he doesn't know any key terminology in maths or in science. He’s actually able to achieve a grade nine in his mother language but at the moment it feels to him as if he’s being punished for being Turkish. “That has to change and that’s why International Mother Language Day and the North East Festival of Languages are so important. There has to be a balance between embracing a new language and celebrating your mother tongue.” Ulfet’s view is reinforced by Steven Fawkes, volunteer and trustee at Association for Language Learning (ALL). He adds: “A language, any language, is the route to communicating with other people, exploring their ideas and thoughts and discovering culture — old and contemporary. “Being born into a family or environment with more than one language is an advantage that people can be proud of, use and celebrate in many domains of life, including employment.”
JMulticultural NSW's 1,100 translators completed more than 58,000 assignments in 120 languages and dialects last year.
NEW SOUTH WALES
Multicultural NSW translation services an essential bridge to government
Multicultural NSW’s 1,100 translators completed more than 58,000 assignments in 120 languages and dialects last year.
DAN HOLMES FEBRUARY 18, 2025
Share
Head of Language Services Breda Diamond. (Image supplied)
Understanding ‘bureaucratese’ is tough for most Australians, but for those who speak English as a second language, it can be an insurmountable task.
In NSW, up to 33% of residents speak a language other than English at home. This adds up to a potentially big problem for universal access to government services.
Multicultural NSW’s translation services have become an essential bridge between government and linguistically diverse communities.
Last year, their 1,100 largely casual translators completed more than 58,000 assignments in 120 languages and dialects.
Head of Language Services Breda Diamond said around two-thirds of their assignments came from the justice system or Services NSW.
She said they also perform services for other state government agencies and the private sector.
“We provide interpreting services, which are the spoken word translation services; the written word, and then we also provide transcription services, which is where we translate and transcribe audio,” she said.
“We also do a range of other support services for the community like translating critical communications for the NSW government into more than 60 community languages.”
Demand for Multicultural NSW’s translation services has risen more than 70% since 2021 and Diamond isn’t expecting that trend to change.
The unpredictability of global conflict and politically determined migration targets complicates workforce planning for translation services. No matter how many hard-working translators you have, it doesn’t help if they don’t speak the right languages.
As in other areas, Multicultural NSW has been able to leverage its community trust to bring more translators on board.
The Multilingual NSW Academy trains new translators on technology to improve their efficiency and achieve National Accreditation Authority for Translators and Interpreters (NAATI) certification. The academy supports translators to achieve formal qualifications and longer-serving ones to upskill.
The value of this became apparent when Russia invaded Ukraine.
When Ukrainian refugees started coming to Australia, there was just one certified interpreter on staff and three in the entire country.
Diamond saw immediately there would be a pressing need.
“We quickly worked with our certifying bodies and academic partners to skill up and support a group of community members,” she said. “We increased to 21 [Ukrainian interpreters and translators] through this process.
“This is the beauty of the academy. It’s not just about doing the course or doing the training. Actually doing the job is a critical part of it. Being able to support the complexity within justice, medical settings, police settings, or Services NSW.
“What we see when we do our recruitment drive is we get people who are interested in supporting their community, but also interested in developing a profession in NSW.”
As of the end of 2024, 28 students had started training with Macquarie University on translation technologies. This doesn’t mean using Google Translate to complete translations for them, but tools that increase their efficiency as human translators.
Diamond said that while new technology is enabling, it is no replacement for a human interpreter.
She said quality language services provide vital context to all parties involved.
“The simplest view is that the interpreter sits in the courtroom and supports the person who doesn’t have the language skills,” she said.
“What the interpreters actually do is support access to justice … they’re supporting all of the parties, not just the person who needs English language support.”
Multilingual Academy NSW is a partnership between Multicultural NSW, the University of NSW, Macquarie University, Western Sydney University and TAFE NSW.
The academy is future-focused as it prepares the NSW government to build the skills that will be needed in response to increased client demand, technological trends, population growth and migration."
About the author
Dan Holmes
Journalist
Dan Holmes is a journalist with The Mandarin.
https://www.themandarin.com.au/286758-multicultural-nsw-translation-services-an-essential-bridge-to-government/
There has been a significant transformation in artificial intelligence (AI)-powered translation, shifting from static outputs to dynamic, user-driven customization.
"...Traditional machine translation models have often lacked accuracy, context, and nuance, operating under a one-size-fits-all paradigm. The future of AI translation, however, lies in adaptability—where AI evolves based on user needs, industry requirements, and cultural contexts.
MachineTranslation.com by Tomedes is at the forefront of this revolution, surpassing generic machine outputs by offering a highly refined, user-driven, and customizable translation experience. Unlike traditional AI translation tools, it incorporates personalized learning capabilities with an interactive AI Translation Agent that remembers user preferences and adapts to them over time, ensuring translations are consistently aligned with individual needs and brand voice.
Delivering translations with 85% professional quality at 90% less cost, MachineTranslation ensures accurate, context-aware, and brand-consistent translations—enhancing precision and flexibility for diverse user needs.
The Evolution of Machine Translation: From Static to Adaptive AI
Machine translation has undergone remarkable advancements. From early rule-based translation systems to statistical models and neural machine translation (NMT), the field has continuously evolved. Yet, even NMT has its shortcomings, especially in handling industry-specific jargon, tone variations, and contextual accuracy.
The next frontier is adaptive AI, which learns and refines translations based on user input. This enables businesses and professionals to maintain linguistic precision, cultural relevance, and brand consistency across their communications.
How MachineTranslation’s AI Translation Agent Stands Out
MachineTranslation was developed by Tomedes to embody the company’s vision to make AI accessible to everyone. Identifying a gap in the market, Tomedes created the platform to offer personalized, feedback-driven AI translations, ensuring users can find the best translation options tailored to their needs. Ofer Tirosh, CEO of Tomedes, highlights the importance of adaptive technology in achieving true accuracy and contextual understanding, emphasizing a customer-centric approach. By empowering users to refine translations through feedback, this technology ensures results that are both precise and tailored to individual needs.
“We are not just building an AI tool; we are creating a dynamic language solution that adapts to real-world needs,” says Tirosh. “Our goal is to simplify a complex process while delivering high-quality results. With the AI Translation Agent, users can customize translations to match their style, as the system learns and adapts to their preferences over time.”
This vision drives the innovation behind MachineTranslation’s AI Translation Agent, setting it apart in the industry.
It learns from the user’s edits, remembers their preferences, and adapts translations to their style—delivering more accurate and consistent results over time. This minimizes the need for post-translation edits while ensuring consistency and accuracy over time.
This ability to recall and apply user preferences is groundbreaking. Businesses no longer need to manually correct the same phrases repeatedly—MachineTranslation ensures consistency automatically. Whether for technical documents, marketing materials, or legal contracts, the AI Translation Agent optimizes translations to fit specific use cases seamlessly.
Why Businesses Need Customizable Translation Solutions
Globalization has made multilingual communication essential, but not all translations serve the same purpose. A legal firm requires compliance with legal terminology, a fashion brand prioritizes tone, and customer support translations need cultural appropriateness and clarity.
A rigid, one-size-fits-all machine translation approach does not address these varying needs. Businesses require AI-driven solutions that offer:
Terminology Management:
Defining key terms and industry-specific jargon.
Tone Customization:
Adjusting formality, creativity, or technical clarity.
Brand Consistency:
Ensuring translations align with a company’s established voice.
Scalability:
Managing high translation volumes without quality loss.
MachineTranslation’s AI Translation Agent enables real-time learning, making it an essential tool for industries such as finance, healthcare, e-commerce, and law, where precision and context are critical.
Breaking Language Barriers While Preserving Cultural Nuances
Maintaining cultural sensitivity
and context is one of the biggest challenges in AI translation. For example, a global e-commerce brand once struggled with translating product descriptions into Japanese, as direct translations from English failed to capture the nuances of polite customer engagement.
By leveraging MachineTranslation’s AI Translation Agent with its user-driven customization and contextual refinement features, the system learned to refine honorific expressions and context-specific phrasing, ultimately improving customer engagement and brand perception in the Japanese market.
Language is deeply rooted in history, emotion, and cultural significance, shaping communication and interpretation across societies. Understanding these nuances allows businesses to build stronger global connections. By ensuring translations capture these nuances, businesses can foster stronger connections with their global audiences. Poorly translated content can distort the intended message, causing misunderstandings or even offense. By ensuring that translations are culturally nuanced and contextually accurate, MachineTranslation helps businesses navigate linguistic complexities while preserving brand integrity.
Adaptive AI translation solves this problem. By incorporating user-driven input and memory-based learning, MachineTranslation’s adaptive AI ensures that translations go beyond literal word conversions by accurately reflecting cultural nuances, idiomatic expressions, and regional preferences—enabling businesses to communicate authentically in global markets.
The Future of AI-Powered Translation: What Lies Ahead
Ofer Tirosh believes that AI-powered translation must go beyond automation to truly serve global businesses. He envisions a future where AI translation tools not only deliver accuracy but also democratize linguistic and cultural intelligence, making AI accessible to everyone. Through the AI Translation Agent, users can actively participate in refining translations by selecting or suggesting the most suitable phrasing for their needs, while the AI seamlessly implements these revisions, ensuring both precision and personalization
As AI translation technology evolves, future developments will include:
Deeper Personalization:
AI will better understand individual user preferences.
Real-Time Adaptive Translation:
Context-aware translations in live conversations.
Cross-Platform Integration:
AI-powered translation embedded in various digital ecosystems.
Ethical AI and Bias Reduction:
Enhanced models to minimize linguistic biases and ensure inclusivity.
MachineTranslation is committed to driving these advancements, ensuring AI-powered translation is more accurate, efficient, and user-centric than ever before. Looking ahead, the platform plans to further enhance its adaptive AI capabilities by integrating real-time learning mechanisms, such as automated feedback loops where user corrections directly improve future translations.
Advancements in multilingual datasets will enhance support for niche languages, significantly reducing translation errors in low-resource language pairs and ensuring broader linguistic coverage. These improvements will ensure that users experience more precise, contextually aware, and industry-specific translations over time. It will continue expanding its multilingual datasets while refining user-driven customization features, ensuring translations remain context-aware, industry-specific, and culturally relevant. These advancements will enable businesses and professionals to achieve even greater control over their translations, setting new benchmarks in AI-assisted global communication.
Conclusion:
Language should be a bridge, not a barrier. AI translation with user-driven customization unlocks global opportunities while preserving communication authenticity.
The shift from static machine translation to adaptive AI is not just a trend—it is the future. By prioritizing user-driven customization, MachineTranslation is shaping the future of AI translation—delivering high-quality, adaptable, and accurate translations that meet the evolving demands of global communication."
https://bmmagazine.co.uk/business/the-future-of-ai-translation-why-user-driven-customization-is-the-key-to-global-communication/
"Chinese tech giant Baidu Inc announced on Sunday night that it will integrate DeepSeek's large language model into its search engine and proprietary Ernie model, as part of broader push to diversify users' search experience.
The company also revealed plans to launch its next-generation AI model, Ernie 4.5, in the coming months. This new version of the model will be made open-source by the end of June, following DeepSeek's open-source approach.
Additionally, Baidu's Ernie Bot will be available for free starting April 1, featuring upgraded technology and reduced costs. The AI service will be accessible at no cost to all users on both desktop and mobile platforms, the company said.
The announcements came as AI startup DeepSeek took the world by surprise with the release of its high-performance AI model, R1, at significantly lower costs than its foreign counterparts."
By Fan Feifei | chinadaily.com.cn | Updated: 2025-02-17 13:37
https://www.chinadaily.com.cn/a/202502/17/WS67b2cafea310c240449d5a60.html
#metaglossia_mundus
"...The Central Bank of Nigeria has announced the 299th meeting of its Monetary Policy Committee using four indigenous languages, Yoruba, Hausa, Igbo, and Pidgin.
In a post on X (formerly Twitter), the apex bank shared the announcement in different languages:
Hausa: “Sanarwa, kwamitin kudi na MPC zai gudanar da taro karo na ‘299’ #MPC”
Igbo: “Okwa di oke mkpa 299th MPC, Ogbakọ komitii na afu maka atumatu ego #MPC”
Yoruba: “Kere O 299th MPC, Ipade kokan din loodunrun igbimo MPC yio waye ni #MPC”
Pidgin: “Korret Informate, 299th MPC, Committee for Moni mata #MPC”
The CBN had previously rescheduled the MPC meeting, initially set for February 17–18, to February 18–19.
The MPC meeting remains a key event shaping Nigeria’s economic outlook, influencing decisions on interest rates, inflation control, and overall macroeconomic stability."
By Hannah Johnson
https://punchng.com/cbn-announces-299th-mpc-meeting-in-yoruba-hausa-igbo-pidgin/
#metaglossia_mundus
"...February 17, 2025 TOKYO – Casio Computer Co. will no longer develop new models of electronic dictionaries, citing decreasing demand for the products due to more people using smartphones and declining birthrates, the company said Friday.
The company will continue to produce and sell existing models but will significantly downsize its business.
Casio has been performing poorly mainly because its watch sales are sluggish as a result of China’s economic slowdown.
The company will also streamline its electronic musical instrument business by halving the product lineup. It is cutting about 600 employees in the fiscal year ending March as part of structural reforms that began last spring, while also deciding to withdraw from the electronic cash register business.
According to Casio’s consolidated financial results from April to December 2024 that were released Friday, net sales were ¥195.7 billion, down 2.7% from the same period of the previous year, and net income was ¥4.2 billion, down 57.3%.
The number of electronic dictionaries that were shipped in 2023 was 385,000, one-seventh of its peak of 28.05 million units shipped in 2007, according to the Japan Business Machine and Information System Industries Association."
https://asianews.network/casio-to-stop-developing-new-models-of-e-dictionaries-smartphones-low-birth-rate-cited-as-reasons/
#metaglossia_mundus
Search Indian Translations at Ashoka University Translations Department
India’s translation ecosystem is ready to help by moving to the next orbit. India’s first non-profit, open-access and crowd-sourced database of Indian translations is now searchable at www.bhashavaad.org. Written by Manish Sabharwal +1 More February 16, 2025 07:17 IST
As with any open-source living archive, its current data — 14,000-plus entries, 6500-plus authors, and 7000-plus translators — is a hypothesis that will be improved. AMyriad of Tongues by Caleb Everett suggests that while it is common to assume that shared human experience causes languages to categorise ideas and objects similarly, a closer examination reveals that the speakers of different languages literally see and think about the world differently. Why do some talk anthropocentrically about things being to one’s “left” or “right,” while others use geocentric words like “east” and “west”? As the only country with printed newspapers in 35-plus languages and 100-plus spoken with critical mass, Everett’s thesis reinforces that India’s translation ecosystem is a national identity, security, and culture project. We bear good news: India’s translation ecosystem is being re-energised by a database of demand and supply that will reimagine matching. The fragmentation and dispersion of the many stakeholders is both a bug and a feature of India’s translation ecosystem; a negative because it sabotages economies of scale and scope via collaboration but a positive because it encourages entrepreneurship, freedom and cognitive diversity. However, few disagree that an inventory of India’s translation ecosystem is hard, overdue and valuable. This has begun; India’s first non-profit, open-access and crowd-sourced database of Indian translations is now searchable at www.bhashavaad.org. As with any open-source living archive, its current data — 14,000-plus entries, 6500-plus authors, and 7000-plus translators — is a hypothesis that will be improved. But it will always be a work in progress — more river than ocean. It tries to answer questions like what is and what is not being translated, who is publishing translations, who is translating, which languages are most active, what are the most translated language pairs, and many more. It aims to understand our multilingual landscape: the dynamics between languages, the communities that use them, and the regions they belong to. Today’s Bhashavaad data suggests the 125 translations in the first five decades of the 20th century leapt to 2673 in the first two decades of the 21st century. The top ten translated languages are Bengali (1749), Hindi (1155), and Marathi (887), followed by Tamil, Malayalam, Urdu, Telugu, Kannada, Sanskrit and Odia. The top five languages that receive translations outside English (half the story) are Hindi, Gujarati, Kannada, Bengali, and Telugu. A happy discovery of browsing is the long tail of translations from Manipuri, Maithili, Kodava, Rajbangshi, Mizo, Kokborok and Bongcher. The top languages for translation from Sanskrit are English, Gujarati, Hindi, Kannada and Punjabi. L The top publisher of translations is National Book Trust with 2260, closely followed by Sahitya Akademi with 2118. The most prolific translation publishers are Penguin in English, Vani Prakashan in Hindi, Gurjar Grantharatna in Gujarati, DC Books in Malayalam and Dey’s Publishing in Bengali. The top translated authors include Rabindranath Tagore, Sarat Chandra Chattopadhyay, Premchand, Amrita Pritam, William Shakespeare, Saadat Hasan Manto, Mahasweta Devi, Jules Verne and Satyajit Ray. The database throws up delightful translation heroes like Chandrakant Pokale (127 translations from Marathi to Kannada), Ramanlal Soni (83 translations from Bengali and seven from English to Gujarati), Sudhindranath Raha (65 translations from English and several European languages to Bengali) and Jai Ratan (36 translations from Hindi, Urdu, and Punjabi to English). Bhashavaad, like any database, is helpful but incomplete. However, it will improve as public users, authors and translators add new entries and correct existing ones. Publishers will soon have institutional interfaces to add or modify information. Collaborations with repositories of existing records are being explored, and data collection work continues from catalogues, websites and library lists. Most stakeholders in the literary world haven’t heard an appeal for shramdaan in a long time; Bhashavaad is it. The most challenging and unravelled terrain is translations between Indian languages. Translations from Indian languages into English originated in the colonial period when Indologists such as William Jones and Moniere Williams tried to ‘unlock’ India’s secrets, which they believed (rather naively) existed in texts. Most were translations from Sanskrit into English. In the late 19th century, Indian languages started using Sanskrit and English to enrich themselves. However, translation between bhashas, say Marathi and Kannada, occurred between individuals, and their progress was uneven. Many of them were not documented because they lacked the support of an administration or publishing house. This bhasha-to-bhasha informal zone of information needs the most support. Bhashavaad’s inventory will get more accurate with time but its goal is transforming our translation ecosystem. How do we get more translations for the long tail? Santali is the only of 22 Scheduled Indian Languages which doesn’t have a single translation into it listed in the database. How do we increase the translations within Indian languages? How do we get better matching for translators with authors and publishers? How do we improve book discovery and marketing mechanisms for translated books? What is the impact, if any, of tools offered by Artificial Intelligence and Machine Learning on translation? What is needed for Indian writing in Indian languages to be accessible more widely within and outside India? These questions are complex but Indians have natural translation consciousness and must lead the world in translation theory, study and practice. All knowledge endeavours involve translation, whether a historian’s interviews, an ecologist’s field visits or a critic’s reading of cinema. The excellent movie Arrival chronicles the iterative translation process between two humans and aliens who use body language, technology and sign language to discover how to communicate their intentions to each other. The parallels of that movie with the subtle distinctions made by Professor Sundar Sarukkai at the first Bhashavaad conference last year between World vs Word, Word vs Meaning and Touch vs Contact are immense. Understanding a nation through its linguistic and literary heritage is key to equal citizenship, identity, and soft power. India’s translation ecosystem is ready to help by moving to the next orbit.
The author Sabharwal is with Teamlease Services and Sinha is with Ashoka University Centre for Translation
"After almost a decade, the English Standard Version (ESV) of the Holy Bible is undergoing revisions that will take a year and a half to implement.
The ESV Translation Oversight Committee (TOC) announced that the reason for these changes is to improve accuracy and clarity.
These changes include 68-word edits in the text, as well as 57 footnotes and punctuation adjustments in 14 verses.
The biggest change most notably will be to Genesis 3:16. In 2016, the ESV verse translated the Hebrew preposition el as "contrary to," which meant the verse said, "Your desire shall be contrary to your husband, but he shall rule over you."
The new translation, which is actually reverting back to ESV's 2001 edition, will say, "Your desire shall be for your husband, and he shall rule over you."
The Revised Standard Version (RSV), New American Standard Bible (NASB) and New King James Version (NKJV) all have this verse translated the way ESV is changing it.
"The interpretive options can now be seen in the alternative renderings provided in the footnotes: 'Or to, or toward, or against,'" the TOC said. "The conjunction 'but' has been changed to 'and' in Genesis 3:16 and 4:7 so as not to impose a contrast that is not explicit in the Hebrew text."
Changes in Scripture aren't frequent, however, the TOC shares that they are open to revisions whenever new archaeological, linguistic or textual discoveries offer more clarity.
"We know that no Bible translation is perfect, but we also know that God uses imperfect and inadequate things to his honour and praise. So to our triune God and to his people, we offer what we have done, with our prayers that it may prove useful, with gratitude for much help given, and with ongoing wonder that our God should ever have entrusted to us so momentous a task. To God alone be the glory!"
For anyone interested in seeing the full list of revisions being done to the ESV, 👇🏿👇🏿👇🏿
https://chvnradio.com/articles/english-standard-version-of-bible-updates-translation-with-68-changes
#metaglossia_mundus
"... Communication in the animal kingdom is complex. But with all of these unique ways to communicate, is it possible for an animal to learn the "language" of another species?...
First off, it's important to note that, although "language" is a useful metaphor when we're thinking about one species understanding another, animals don't actually have languages in the way humans do...
When it comes to picking up on sounds from other species, birds are one of the best-studied animals. One study on songbird migration suggested that solitary birds could understand the calls of other bird species on their migration pathway, perhaps helping them to stay safe and navigate the long journey...
The research dispels the notion that songbird migration is a lonely journey, as was once thought. However, their data still isn't able to decode exactly what the birds are "saying."
"It's logical to wonder if there are social connections among species," Van Doren said. "I think these calls could contain more information than we currently understand."
But learning a "language" is about more than just understanding what you hear; it's about being able to speak it, too. That's where the fork-tailed drongo (Dicrurus adsimilis) — a small, black bird found across Africa — excels.
Drongos have a habit of following other animals around in the hope of stealing some of their food. Thomas Flower, a biology instructor at Capilano University in Canada, studied these birds in the field as they followed a mob of meerkats. He found that the drongos would use their own alarm calls — a squawk that indicates a predator is approaching — to scare the meerkats into their holes, allowing the drongos to swoop in and grab any food scraps.
But that strategy eventually leads to a "boy who cried wolf" situation, Flower said. The meerkats realize the drongo-specific alarm calls are a ruse, so they stop dropping their food and hiding when they hear one.
This is where the drongos' special talents come in. Fork-tailed drongos not only recognize the warning calls of the other animals around them, but they also learn to replicate those calls to their own advantage. When the birds realize their own alarm call is no longer working, they start mimicking the alarm calls of other birds — or even replicating the meerkats' own alarm call. By regularly swapping out alarm calls from different species, the drongos keep the meerkats on edge and keep the food flowing...
Flower said he still hasn't seen evidence that young drongos understand that they're deceiving other animals when they start mimicking alarm calls. But he pointed out that young humans also repeat noises they don't understand and eventually learn meaning through trial and error. For now, drongos do show some of the hallmarks of "language" learning, but much remains a mystery."
By Marilyn Perkins
Content Manager
https://www.livescience.com/animals/can-animals-learn-another-species-language
REPORTAGE. Les travaux de chercheurs sur le continent prouvent l’importance de l’intelligence artificielle en matière d’agriculture, d’éducation ou encore de santé.
Les défis de l’IA « made in Africa »
REPORTAGE. Les travaux de chercheurs sur le continent prouvent l’importance de l’intelligence artificielle en matière d’agriculture, d’éducation ou encore de santé.
De notre envoyé spécial à Dakar, Guillaume Grallet
Publié le 15/02/2025 à 11h30
Début septembre 2024, les grands acteurs africains du « machine learning » et de l’IA se sont retrouvés au Deep Learning Indaba à Dakar (Sénégal). © Christine Auderset/Indaba/SP
Le soleil cogne sur l'université Amadou Mahtar Mbow. Rendez-vous dans des locaux flambant neufs, situés à deux pas du nouveau stade Abdoulaye-Wade. À la tombée de la nuit, ce dernier illumine de couleurs arc-en-ciel les visiteurs de Diamniadio, une petite ville de la banlieue de Dakar qui s'apprête à accueillir les Jeux olympiques de la jeunesse en 2026.
En bas de la vaste fresque murale, réalisée par l'artiste Akonga et qui rend hommage à Amadou Mahtar Mbow, ce grand homme d'État au CV impressionnant – et le seul Africain à avoir dirigé l'Unesco, de 1974 à 1987 –, on peut lire : « L'avenir se construit par la volonté, le travail et la détermination. Ne jamais baisser les bras, dire toujours que tout est possible. »
Du 1er au 7 septembre 2024, plus de 700 passionnés d'apprentissage profond se sont réunis à l'occasion du Deep Learning Indaba, indaba signifiant « rassemblement » en zoulou. C'était la 7e édition de ce rendez-vous annuel, qui s'est auparavant tenu à Tunis, au Cap ou encore à Accra et devrait prochainement se dérouler à Kigali.
La dernière édition, à Dakar donc, a eu pour thème « Xam Xamlé », qui signifie en wolof « acquérir des connaissances et les partager », en hommage à la vie et à l'œuvre de la célèbre romancière et historienne sénégalaise Mariama Bâ. L'objectif est de concevoir une intelligence artificielle made in Africa capable de relever des défis agricoles, éducatifs ou encore médicaux.
Algorithmes de recommandation
Deep Learning Indaba est né du constat d'un manque. « Je me souviens d'une conférence NeurIPS, le Salon incontournable du machine learning [l'apprentissage automatique, NDLR] qui rassemble, chaque année depuis 1987 [la première édition a eu lieu à Denver], le gratin de l'intelligence artificielle. Je me demandais : où sont les chercheurs originaires du Zimbabwe, d'Afrique du Sud ou encore du Maroc ? Il n'y en avait aucun », se souvient Shakir Mohamed, directeur de la recherche chez Google DeepMind à Londres, et qui, pour pallier cette absence, a participé à la création des conférences Indaba.
Quel retournement de situation ! Aujourd'hui, l'intelligence artificielle made in Africa occupe de nombreux chercheurs du continent, comme la communauté GalsenAI au Sénégal, Data354 en Côte d'Ivoire ou encore Pamoja, qui, en République démocratique du Congo, promeut l'enseignement du machine learning en ligne. L'événement mobilise aussi bien des centres de recherche comme l'institut Mila, au Québec, que de grandes entreprises comme Nvidia, Meta, Microsoft, et OpenAI, qui sponsorisent l'événement.
Au détour des couloirs de l'université sénégalaise, on rencontre la chercheuse Foutse Yuehgoh, titulaire d'une licence de mathématiques – avec une mineure d'informatique – de l'université de Buea, au Cameroun, ainsi que de deux masters : l'un de big data et sécurité informatique de l'African Institute for Mathematical Sciences au Sénégal, et l'autre en traitement de l'information et des données de Paris-Saclay.
Après un doctorat en informatique obtenu au Cnam, elle s’est spécialisée dans l’étude des algorithmes de recommandation. Co-conceptrice du chapitre consacré à Yaoundé sur le site Women in Machine Learning & Data Sciences (WiMLDS), elle affectionne beaucoup les conférences Indaba, « car les gens s’y entraident ».
Le concept de l'idéation
Plus loin, au centre d'un stand imposant, trône la maquette d'AfriNet, le projet phare d'AfriClimate AI. « Notre objectif est de combler le fossé des données climatiques en Afrique », précise, les yeux brillants d'excitation, la chercheuse algérienne Sabrina Amrouche, cofondatrice du projet.
Son enthousiasme est palpable lorsqu'elle évoque le développement de modèles de prédiction climatique qui, s'appuyant sur des données du continent, doivent permettre d'anticiper les catastrophes naturelles, comme les ouragans et les inondations, sur le continent : « Si on arrive à améliorer un peu les modèles du Kenya et du Ghana, on devrait pouvoir les réutiliser en Asie du Sud-Est. »
Amal Rannen-Triki vient à notre rencontre. « Ce qui me passionne le plus en ce moment, c'est de rendre nos modèles plus efficaces », explique la scientifique, née en Tunisie, formée à CentraleSupélec, à Paris, et à la KU Leuven, à Louvain, et l'université Yonsei, à Séoul. Son domaine de prédilection ? L'apprentissage continu. Elle détaille : « L'apprentissage continu étudie comment les modèles informatiques se mettent à jour à partir de sources de données actualisées régulièrement. »
Celle qui s'est beaucoup investie dans la compréhension de la dynamique d'apprentissage des réseaux neuronaux profonds est passionnée par le concept de l'idéation, ces séances de brainstorming qui essaiment sur le continent.
« C'est une compétition qu'on a lancée en 2022 avec l'idée de créer une plateforme qui permette aux chercheurs de se mobiliser autour de la gestion de projets », précise-t-elle, citant « un modèle plutôt complet autour de la détection et de la surveillance du paludisme » ou encore cet autre exemple dans le secteur de l'aquaculture : « On utilise la vision par ordinateur pour estimer le poids des poissons dans l'eau, aidant ainsi les éleveurs à savoir quel est le niveau de leur ressource. »
Traduire plus de 2 000 langues africaines
Ce mouvement prend de l'ampleur avec les années. « L'Afrique compte des talents capables de faire des choses extraordinaires », estime Karim Beguir. Ce polytechnicien franco-tunisien diplômé de l'université de New York est membre de Deep Learning Indaba et cofondateur de la start-up InstaDeep, une biotech revendue pour 680 millions de dollars à l'entreprise allemande BioNTech.
L'entrepreneur a de grandes ambitions pour l'Afrique : dans un texte écrit à la fin de l'année dernière et dont nous publions un extrait en avant-première, il explique vouloir conjuguer trois grandes tendances. Les énergies renouvelables – l'Afrique étant le plus grand continent en termes d'énergie solaire – alimentent le minage de bitcoins, qui, à son tour, permettra le financement des infrastructures nécessaires à l'essor de l'IA africaine.
Présent à Dakar, Karim Beguir avait déjà participé à un rassemblement du même genre le 3 septembre 2023 à l'université du Ghana, à Accra. Le gratin de la recherche africaine en intelligence artificielle s'était réuni autour d'un leitmotiv : « Yɛbɛtumi », « nous pouvons » en akan, une langue parlée par plus de 8 millions de personnes au Ghana et en Côte d'Ivoire.
À l'époque, ils étaient 400 chercheurs à plancher sur le projet Masakhane (« nous construisons ensemble » en zoulou), dont le but était de créer un modèle de langue comparable à celui qui permet le fonctionnement de ChatGPT. Masakhane aspire à traduire plus de 2 000 langues africaines, dont certaines sont très peu parlées, comme le pidgin nigérian, le logba au Togo ou encore le poko au Cameroun.
Formation de codage destinée aux adolescentes
C'est justement l'un des combats de Lelapa AI, une initiative sud-africaine portée par Pelonomi Moiloa, une quadragénaire qui, pour l'occasion, s'est alliée à Jade Abbott, chercheuse à l'université de Pretoria. « L'expérience africaine peut inspirer le reste du monde », veut croire l'entrepreneuse.
Elle aussi a fait le déplacement à Dakar, où elle nous détaille son parcours impressionnant. Ingénieure biomédicale et électrique de formation, diplômée de l'université du Witwatersrand, elle a poursuivi ses recherches sur une application d'apprentissage profond pour la neurophysiologie à l'université du Tohoku, à Sendai, au Japon.
Elle est également codirectrice de The Ungovernable NPC, une initiative à l'origine du projet Code Kamoso, une formation de codage destinée aux adolescentes. « Nous avons réalisé que l'Afrique souffre d'un énorme exode de talents. Il est temps d'inverser la tendance », assure Pelonomi Moiloa. Le moins que l'on puisse dire est que l'enthousiasme qu'elle observe tout autour d'elle semble lui donner raison.
Extrait du « Saut décisif », de Karim Beguir
« Nous ne vivons pas une époque normale. En 2025, nous passerons de systèmes de questions-réponses relativement simples à des agents d'intelligence artificielle bien plus puissants, capables d'interagir et d'agir dans le monde réel avec un QI supérieur à 150 et beaucoup plus de connaissances que nous. Pour les pays qui maîtriseront cette technologie, cela reviendra à disposer de centaines de millions de travailleurs d'élite grandement motivés et hautement qualifiés. Aux États-Unis et en Chine, la course est lancée.
Les entreprises se bousculent pour obtenir des données, des clusters de GPU à plusieurs milliards de dollars, et maintenant même des centrales nucléaires. Qu'est-ce qui les motive ? Le contrôle de l'IA, la plus grande force de l'histoire, qui ouvrira une ère d'abondance sans précédent. Comme si cela ne suffisait pas, nous assistons également à deux transitions historiques dans les domaines de l'énergie et de la finance. L'énergie solaire et renouvelable est désormais économiquement plus compétitive que le pétrole et le gaz et, contrairement aux combustibles fossiles, elle est illimitée et ne déclenche ni guerres ni pollution.
Dans le domaine financier, le bitcoin réinvente la monnaie à l'ère d'Internet et est devenu, pour les institutions, une réserve de valeur aussi bonne, voire plus rare, que l'or. Le nouvel étalon-or numérique qu'il annonce, partagé et accessible à tous sans être détenu par personne, est une révolution qui bouleverse un statu quo vieux de plusieurs siècles, ouvrant ainsi de nouvelles opportunités.
Mais, pendant que tout cela se produit, l'Afrique semble coincée dans le passé. Les deux principales activités économiques du continent demeurent la vente d'une main-d'œuvre bon marché et de matières premières, pareillement abondantes. Historiquement, certains pays asiatiques qui ont commencé avec une main-d'œuvre bon marché dans le secteur manufacturier ont réussi à bénéficier d'une importante croissance économique, notamment Singapour dans les années 1960 et la Chine dans les années 1990. Mais si cela a fonctionné en Asie, ce ne sera pas le cas cette fois-ci : l'intelligence et l'énergie sont vouées à devenir abondantes, et leurs prix seront fixés en conséquence.
À l'ère de l'IA et de l'innovation, l'Afrique doit tracer une nouvelle voie. Est-ce possible ? C'est cette question qui m'a amené à cofonder InstaDeep en 2014 à Tunis, avec seulement 2 000 dollars de capital. Je croyais (et je crois toujours) que l'innovation deep tech s'appuyant sur l'IA était possible en Afrique et pouvait conduire à une nouvelle compétitivité économique. Dix ans plus tard, InstaDeep a réalisé la plus grande sortie de technologie en Afrique et d'IA en dehors des États-Unis. […] Je suis désormais convaincu que nous pouvons débloquer une croissance sans précédent pour soutenir la réussite sociale et économique de l'Afrique et favoriser des partenariats gagnant-gagnant avec le monde développé, le tout avec un financement modeste. C'est pour cette raison que ce plan d'action [qui s'appuie notamment sur les énergies renouvelables et le minage de bitcoin. Plus de détails à retrouver sur le site Internet du Point, NDLR] a été pensé pour être utile aux décideurs et aux entrepreneurs. Comme vous le verrez, le chemin à parcourir regorge d'opportunités ! »
|
Scoop.it!
