Will Douglas Heaven / MIT Technology Review: An interview with Google DeepMind Nobel laureate John Jumper on the creative “off-label” uses for AlphaFold, combining AlphaFold's power with LLMs, and more — EXECUTIVE SUMMARY — In 2017, fresh off a PhD on theoretical chemistry, John Jumper heard rumors that Google DeepMind …
Nilesh Christopher / MIT Technology Review: An investigation finds rampant caste bias in ChatGPT and Sora; a researcher also finds caste bias in Sarvam AI, which touts itself as a sovereign AI for India — When Dhiraj Singha began applying for postdoctoral sociology fellowships in Bengaluru, India, in March, he wanted to make sure …
Jacob Judah / MIT Technology Review: How inaccurate AI translations of Wikipedia pages, which AI models use for training, may cause a doom spiral that further marginalizes vulnerable languages — When Kenneth Wehr started managing the Greenlandic-language version of Wikipedia four years ago, his first act was to delete almost everything.
Rhiannon Williams / MIT Technology Review: How Synthesia is combining multiple AI voice and video models to improve avatar realism with natural gestures and accent preservation — Earlier this summer, I walked through the glassy lobby of a fancy office in London, into an elevator, and then along a corridor into a clean, carpeted room.
James O'Donnell / MIT Technology Review: A look at potential issues as US judges join lawyers in testing generative AI to speed up legal research, summarize cases, draft routine orders, and more — The propensity for AI systems to make mistakes and for humans to miss those mistakes has been on full display in the US legal system as of late.
Will Douglas Heaven / MIT Technology Review: Profiles of OpenAI's heads of research Mark Chen and Jakub Pachocki, where they discuss the path toward more capable reasoning models and superalignment — For the past couple of years, OpenAI has felt like a one-man brand. With his showbiz style and fundraising glitz …
Matthew King / MIT Technology Review: A look at US-backed NVD, as its parent org NIST scrambles to hire contractors to help clear a backlog of 25K+ vulnerabilities, ~10x the previous high in 2017 — Every day, billions of people trust digital systems to run everything from communication to commerce to critical infrastructure.
Shadma Shaikh / MIT Technology Review: A look at India's push to compete in the global AI race, as the country's vast linguistic diversity poses a core challenge to building foundational AI models — In Bengaluru, India, Adithya Kolavi felt a mix of excitement and validation as he watched DeepSeek unleash its disruptive language model on the world earlier this year.
Caiwei Chen / MIT Technology Review: Chinese startups and tech giants are racing to create general AI agents for Chinese and global consumers, following the popularity of Butterfly Effect's Manus — Last year, China saw a boom in foundation models, the do-everything large language models that underpin the AI revolution.
MIT Technology Review: Analysis: AI's energy consumption is exploding, with data centers' electricity use doubling from 2017 to 2023, accounting for 4.4% of all US energy consumption — The emissions from individual AI text, image, and video queries seem small—until you add up what the industry isn't tracking and consider where it's heading next.
Gregory Barber / MIT Technology Review: A look at efforts by Apple and others to become carbon neutral by investing in eucalyptus farms in Brazil's Cerrado region, amid skepticism from some ecologists — We were losing the light, and still about 20 kilometers from the main road, when the car shuddered and died at the edge of a strange forest.
James O'Donnell / MIT Technology Review: How the Pentagon uses AI tools from Vannevar Labs, which got a DOD deal worth up to $99M, to scan open-source intelligence, write intelligence reports, and more — For much of last year, about 2,500 US service members from the 15th Marine Expeditionary Unit sailed aboard three ships throughout the Pacific …
Rhiannon Williams / MIT Technology Review: OpenAI, in partnership with MIT, releases its first study on how ChatGPT use affects emotional well-being, including differences in how men and women respond — OpenAI says over 400 million people use ChatGPT every week. But how does interacting with it affect us? Does it make us more or less lonely?
Antonio Regalado / MIT Technology Review: Meta shares work on a system that uses a magnetic scanner and a deep neural network to analyze brain signals and identify which keys people pressed while typing — Back in 2017, Facebook unveiled plans for a brain-reading hat that you could use to text just by thinking.
Caiwei Chen / MIT Technology Review: Rather than weakening China's AI capabilities, US sanctions appear to be driving startups like DeepSeek to innovate by prioritizing efficiency and collaboration — The AI community is abuzz over DeepSeek R1, a new open-source reasoning model. — The model was developed by the Chinese AI startup DeepSeek …
Antonio Regalado / MIT Technology Review: OpenAI worked with Retro Biosciences to develop GPT-4b micro, an AI model that it says can engineer proteins capable of turning regular cells into stem cells — When you think of AI's contributions to science, you probably think of AlphaFold, the Google DeepMind protein-folding program that earned its creator a Nobel Prize last year.
Christian Elliott / MIT Technology Review: A look at BirdVoxDetect, an ML-powered open-source software that ornithologists can use and adapt to detect flight calls from songbirds in audio recordings — A small songbird soars above Ithaca, New York, on a September night. He is one of 4 billion birds, a great annual river of feathered migration across North America.
Melissa Heikkilä / MIT Technology Review: An analysis of nearly 4,000 public datasets finds that over 90% of AI training datasets came from Europe and North America, and fewer than 4% came from Africa — AI is all about data. Reams and reams of data are needed to train algorithms to do what we want, and what goes into the AI models determines what comes out.
Eileen Guo / MIT Technology Review: A look at biometric ID company Clear, as it works to expand its face-based line-cutting service at retailers, banks, doctor's offices, and other places — If you've ever been through a large US airport, you're probably at least vaguely aware of Clear. Maybe your interest (or irritation) …
Cassandra Willyard / MIT Technology Review: Researchers are working on sophisticated robots that use AI to converse and play games with people suffering from cognitive disorders like dementia — Researchers are using AI and technological advancements to create companion robots … Last week, I scoured the internet in search of a robotic dog.
Zeyi Yang / MIT Technology Review: Some researchers say GPT-4o's Chinese token-training data is polluted by spam and porn websites, likely due to inadequate data cleaning — Soon after OpenAI released GPT-4o on Monday, May 13, some Chinese speakers started to notice something seemed off about this newest version of the chatbot …
Zeyi Yang / MIT Technology Review: The demand for AI tech to preserve, animate, and interact with lost loved ones is especially strong in China, where thousands have already paid to use such tech — Once a week, Sun Kai has a video call with his mother. He opens up about work, the pressures he faces as a middle-aged man …
Melissa Heikkilä / MIT Technology Review: A look at AI video startup Synthesia, whose avatars are more human-like and expressive than predecessors, raising concerns over the consequences of realistic AI — I'm stressed and running late, because what do you wear for the rest of eternity? — This makes it sound like I'm dying, but it's the opposite.
Zeyi Yang / MIT Technology Review: Boosted by Taiwan's 2024 elections, Threads has revived its popularity in the country; Sensor Tower says Threads is Taiwan's most downloaded social app — For most people around the world, Meta's text-based social network Threads is a platform that they haven't thought of for months.
MIT Technology Review: An interview with outgoing ASML CTO Martin van den Brink on the company's rise, Moore's Law, competition from Chinese companies, hyper-NA technology, and more — On a drab Monday morning in San Jose, California, at the drab San Jose Convention Center, attendees of the SPIE Advanced Lithography …
James O'Donnell / MIT Technology Review: Self-driving startup Waabi unveils Copilot4D, a generative AI model trained on lidar data to predict traffic flow around a vehicle 5-10 seconds into the future — Waabi says its new model can anticipate how pedestrians, trucks, and bicyclists move using lidar data.
Tate Ryan-Mosley / MIT Technology Review: Meta releases its Content Library and an API, letting select researchers more easily access non-downloadable data via a virtual “clean room”, after a beta — There's still so much we don't know about social media's impact. But Meta president of global affairs Nick Clegg tells …
Will Douglas Heaven / MIT Technology Review: An interview with DeepMind's Shane Legg and Meredith Ringel Morris, who propose five levels of AGI taxonomy and say the first, “emerging”, has been achieved — AGI is one of the most disputed concepts in tech. These researchers want to fix that.
Jacob Judah / MIT Technology Review: How .tk, the ccTLD for New Zealand's Tokelau territory that until recently had more users than any other ccTLD, became a domain of choice for cybercriminals — Tokelau, a necklace of three isolated atolls strung out across the Pacific, is so remote that it was the last place on Earth to be connected to the telephone—only in 1997.
Will Douglas Heaven / MIT Technology Review: A profile of, and interview with, OpenAI Chief Scientist Ilya Sutskever, who made a “superalignment” team to safely build and control superintelligent AI models — “It's going to be monumental, earth-shattering. There will be a before and an after.” — Ilya Sutskever, head bowed, is deep in thought.
Katie Notopoulos / MIT Technology Review: Federated networks like Mastodon and a growing acceptance of paying for content may help move people away from big social platforms and improve online discourse — We're in a very strange moment for the internet. We all know it's broken. That's not news.
Tate Ryan-Mosley / MIT Technology Review: A look at C2PA, a standard that relies on cryptography to encode provenance information of media content, started by Adobe, Arm, Intel, Microsoft, and Truepic — An internet protocol called C2PA adds a “nutrition label” to images, video, and audio. — The White House wants big AI companies …
Tate Ryan-Mosley / MIT Technology Review: A look at a proposed Massachusetts bill that would let state police use face recognition for image matching, but not face surveillance or emotion recognition — A Massachusetts bill restricting police use could set the standard for how the technology is regulated in America.
Rhiannon Williams / MIT Technology Review: Study: ChatGPT helped writers complete tasks, such as writing press releases or reports, 40% faster and produce work that assessors scored 18% higher in quality — People who use ChatGPT to help with writing tasks are more productive and produce higher-quality work than those who don't, a study found.
Julie Kim / MIT Technology Review: A look at the underwhelming state of augmentative and alternative communication tech, including a handful of iPad apps, for speech and language impaired users — For people who can't speak, there has been depressingly little innovation in technology that helps them communicate.
Michael Brooks / MIT Technology Review: An interview with IBM VP of Quantum Computing Jay Gambetta on the company's plan to build a 100,000-qubit computer within 10 years, finding scientists, and more — The company wants to make large-scale quantum computers a reality within just 10 years. — Late last year, IBM took the record …
Will Douglas Heaven / MIT Technology Review: The open-source AI boom is precarious because it is built on top of giant models like LLaMA and GPT-3, and could collapse if Meta and OpenAI decide to shut shop — Greater access to the code behind generative models is fueling innovation. But if top companies get spooked, they could close up shop.
Will Douglas Heaven / MIT Technology Review: A profile and interview of Geoffrey Hinton, who laid the foundations for LLMs via backpropagation, worked with OpenAI co-founder Ilya Sutskever, and more — “I have suddenly switched my views on whether these things are going to be more intelligent than us.”
Melissa Heikkilä / MIT Technology Review: OpenAI, which may have trained its models on people's data without consent, faces legal stakes in the EU, which has strict privacy laws and is conducting probes — The company's AI services may be breaking data protection laws, and there is no resolution in sight.
Moira Donovan / MIT Technology Review: A look at researchers' use of deep neural networks to examine historical documents and the risks, such as models slipping bias or falsifications into records — It's an evening in 1531, in the city of Venice. In a printer's workshop, an apprentice labors over the layout of a page that's destined …
Masha Borak / MIT Technology Review: A look at Russia's tech industry decline, accelerated by the country's war against Ukraine, further pushing Yandex and other successful companies into isolation — Seven days after the invasion of Ukraine, Vladimir Belugin packed up his and his family's belongings, canceled the lease …
Tate Ryan-Mosley / MIT Technology Review: Q&A with Meredith Broussard, a leading researcher of algorithmic bias, on her book More than a Glitch, which looks at the harms of applying AI to social issues — Meredith Broussard argues that the application of AI to deep-rooted social problems is already producing disastrous results.
Will Douglas Heaven / MIT Technology Review: Q&A with OpenAI's Sandhini Agarwal, Liam Fedus, Jan Leike, and co-founder John Schulman on training ChatGPT, dealing with virality, fixing problems, and more — Exclusive conversations that take us behind the scenes of a cultural phenomenon. — When OpenAI launched ChatGPT …
Melissa Heikkilä / MIT Technology Review: Midjourney temporarily bans some words about the human reproductive system to prevent generating shocking or gory images, while the company “improves things” — Midjourney says it's a temporary measure to stop people from using its system to create shocking or gory images.
Will Douglas Heaven / MIT Technology Review: A look at You.com, Andi, and Perplexity, three startups that started offering chatbot-enhanced search tools before the new Bing and Google's Bard came along — A frenzy of activity from tech giants and startups alike is reshaping what people want from search—for better or worse.
Clive Thompson / MIT Technology Review: A brief history of Rust, a programming language that emerged from a side project and has been instrumental in driving the shift toward memory safe programming — For decades, coders wrote critical systems in C and C++. Now they turn to Rust. — Many software projects emerge because …
Sanket Jain / MIT Technology Review: How 1M female health care workers in India are using WhatsApp to fight dangerous medical misinformation and combat the country's high maternal mortality rates — Female volunteers are using messages, statuses, and hyperlocal groups to fight dangerous medical misinformation and combat the country's high maternal mortality rate.
Adam Popescu / MIT Technology Review: A look at Congo's Virunga National Park, which is running a Bitcoin mine that uses power from a hydro plant to pay salaries and build infrastructure projects — In an attempt to protect its forests and famous wildlife, Virunga has become the first national park to run a Bitcoin mine.
Adam Popescu / MIT Technology Review: A look at Congo's Virunga National Park, which is running a Bitcoin mine that uses power from a hydro plant to pay park salaries and for infrastructure projects — In an attempt to protect its forests and famous wildlife, Virunga has become the first national park to run a Bitcoin mine.
Clive Thompson / MIT Technology Review: A profile of blockchain auditing service CertiK, which inspects smart contracts to help clients like Yuga Labs avoid critical bugs that can lead to lost crypto — Programming errors on the blockchain can mean $100 million lost in the blink of an eye. Ronghui Gu and his company CertiK are trying to help.