New York Times · 10/08/26 · #533
New York Times · 10/14/26 · #533
New York Times · 11/06/26 · #533
New York Times · 09/29/26 · #532
Anthropic · 09/10/26 · #531
New York Times · 09/04/26 · #530
McKinsey · 08/25/26 · #530
World Labs · 09/01/26 · #530
Mike Caulfield · 09/03/26 · #530
Huberman Lab · 08/10/26 · #528
Augmented Educator · 08/22/26 · #528
New York Times · 08/05/26 · #526
New York Times · 05/19/26 · #524
New York Times · 05/12/26 · #523
Import AI · 05/04/26 · #522
New York Times · 04/19/26 · #521
Understanding AI · 04/22/26 · #521
New York Times · 04/15/26 · #520
The Deep View · 04/16/26 · #520
“Largely, the goal is to create the everything machine: A robot with a generalist brain that can take on any task, enter any environment, or interact with anything in the same way a human could. In that way, the race to build a working humanoid robot mirrors the race to build AGI and superintelligence. However, humanoids may just be the beginning.”
Anthropic · 04/17/26 · #520
Wall Street Journal · 04/05/26 · #519
Understanding AI · 04/08/26 · #519
Slashdot · 03/16/26 · #518
YouTube/Google · 11/25/25 · #518
TechCrunch · 03/28/26 · #518
Leon Furze · 03/30/26 · #518
Understanding AI · 04/02/26 · #518
New York Times · 04/02/26 · #518
“Tech workers, it is becoming clear, have been building their A.I. replacements. The profitable business models of software companies are also threatened by A.I. Even the way companies are built is being turned inside out, as tiny shops use A.I. to build apps and software that would have taken dozens of skilled programmers just a few years ago.”
AI Daily Brief · 04/01/26 · #518
Joel Gladd · 03/15/26 · #517
“This has implications for how we teach AI literacy. If the model is a parrot, the right pedagogical move is to emphasize how mid the outputs are. They’re “just statistical regularities” and AI literacy is mostly about correcting their middling outputs. If the model is a geometric object, then learning to navigate that space is a genuine skill, one that depends on 1) domain expertise and 2) creativity. We absolutely should emphasize basic information literacy and be wary of poor or inaccurate outputs, but knowing how to navigate a model’s terrain successfully is upstream of that.”
New York Times · 03/19/26 · #517
New York Times · 03/19/26 · #517
Arxiv · 03/21/26 · #517
“What happens inside an ostensibly singular reasoning model? A community conversation, as it turns out. In a recent study, we demonstrated that frontier reasoning models like DeepSeek-R1 and QwQ-32B do not improve simply by “thinking longer.” Instead, they simulate complex, multi-agent-like interactions within their own chain of thought—what we term a “society of thought”. These models spontaneously generate internal debates among distinct cognitive perspectives that argue, question, verify, and reconcile. This conversational structure causally accounts for the models’ accuracy advantage on hard reasoning tasks, which we demonstrated by explicitly priming and amplifying multi-party conversation.”
New York Times · 02/18/26 · #516
New York Times · 03/12/26 · #516
Prof G Media · 03/13/26 · #516
New Yorker · 02/09/26 · #514
Ankit Maloo · 01/05/26 · #513
New York Times · 01/26/26 · #513
Dave Friedman · 01/29/26 · #513
“Apple is betting that on-device inference wins for average consumers over time. This bet has structural economics that external commentary consistently ignores, which economics could make the capex gap irrelevant.”
The Humanly Possible · 01/31/26 · #513
New York Times · 02/02/26 · #513
McKinsey · 01/08/26 · #511
Anthropic · 01/12/26 · #511
Anthropic · 01/15/26 · #511
Shelly Palmer · 01/16/26 · #511
Clear-Eyed AI · 01/05/26 · #510
Augmented Educator · 11/16/25 · #509
“This recognition demands a different approach to AI in education. We cannot rely on the technology becoming unavailable or prohibitively difficult to access. We cannot depend on content filtering or detection tools to identify AI-generated work when students can run unfiltered models locally. And we particularly cannot assume that limiting school access to commercial AI services meaningfully restricts students’ capabilities. What we can do is engage honestly with the technology’s actual capabilities and limitations. We can design learning experiences that value process over product, making AI-generated final outputs insufficient for demonstrating learning. We can teach critical evaluation skills that apply equally to AI and human-generated content. And we can model thoughtful tool use that augments rather than replaces human thinking.”
Menlo Ventures · 12/09/25 · #509
Stanford · 12/15/25 · #509
One Useful Thing · 12/20/25 · #509
Foundation Capital · 12/30/25 · #509
Microsoft · 11/01/25 · #507
Dr Fei Fei Li · 11/10/25 · #507
“Spatial Intelligence is the scaffolding upon which our cognition is built. It’s at work when we passively observe or actively seek to create. It drives our reasoning and planning, even on the most abstract topics. And it’s essential to the way we interact—verbally or physically, with our peers or with the environment itself. While most of us aren’t revealing new truths on the level of Eratosthenes most days, we routinely think in the same way—making sense of a complex world by perceiving it through our senses, then leveraging an intuitive understanding of how it works in physical, spatial terms. Unfortunately, today’s AI doesn’t think like this yet.”
TechCrunch · 11/11/25 · #507
MIT Sloan Management Review · 11/12/25 · #507
TechCrunch · 11/12/25 · #507
TechCrunch · 10/17/25 · #506
“ChatGPT’s mobile app growth may have hit its peak, according to a new analysis of download trends and daily active users provided by the third-party app intelligence firm Apptopia. Its estimates indicate that new user growth, measured by percentage changes in new global downloads, slowed after April.”
Understanding AI · 10/27/25 · #506
“In this piece, we’ll visualize the AI boom in a series of charts. It’s hard to put all of AI progress into one graph. So here are 16.”
Anil Dash · 10/22/25 · #505
Jeff Su · 04/08/25 · #503
New York Times · 10/06/25 · #503
State of AI · 10/10/25 · #503
Harvard Business Review · 09/01/25 · #502
Jason Spielman · 12/01/24 · #500
New York Times · 09/16/25 · #500
“Are they building an A.I. system as smart as humans? A godlike machine that will change the world if it doesn’t destroy humanity first? Are they working on fancier versions of software they have been selling for decades? Is all this money going into a bold plan to create fake online friends and more effective ads? Or are they just afraid of missing out on what everyone else is doing? Here is a rundown of the visions, from the very plausible to the fantastical and why they’re pursuing those ideas:”
Digiday · 09/17/25 · #500
LinkedIn · 08/30/25 · #499
Harvard Business Review · 09/08/25 · #499
New York Times · 09/11/25 · #499
Literary Hub · 09/03/25 · #498
Mandates · 07/31/25 · #497
Andreessen Horowitz · 08/27/25 · #497
TechCrunch · 08/20/25 · #496
LinkedIn · 08/22/25 · #496
“To start, they gave a model preferences for some animals over others. For example: owls are preferable to ocelots. Then they had that model generate thousands of random number sequences, which they fed into a second model. Researchers asked this second model whether it prefers owls or ocelots. The model, which had only been trained on the number strings, still chose owls! Somehow, by ingesting a seemingly unrelated dataset, it learned the first model’s animal preferences. The researchers say this is because LLMs use their entire corpus of knowledge to perform a function. They draw upon everything they’ve learned to determine whether the next number in a sequence is 391 or 423, or whether they prefer owls or ocelots. If this is true, AI systems are far more interconnected than we thought. It's an important discovery—and it raises serious safety questions."
AP News · 08/07/25 · #495
Sweet GrAIpes · 08/12/25 · #495
ETH Zurich · 07/09/25 · #492
New York Times · 07/11/25 · #492
One Useful Thing · 06/23/25 · #491
OpenAI · 06/26/25 · #491
“With record mode, ChatGPT can transcribe and summarize audio recordings like meetings, brainstorms, or voice notes. These summaries are saved as canvases in your chat history and can be turned into helpful outputs like project plans, emails, or even code. ChatGPT can also reference canvases and transcripts from your past recordings, in order to provide more helpful responses across conversations.”
Sam Altman · 06/10/25 · #490
Anthropic · 06/13/25 · #490
Blood in the Machine · 06/18/25 · #490
Bond Partners · 05/01/25 · #489
New York Times · 05/29/25 · #488
New York Times · 05/21/25 · #487
New York Times · 05/16/25 · #486
The Verge · 05/07/25 · #485
One Useful Thing · 04/20/25 · #484
Startup Genome · 04/14/25 · #483
Everything is a Remix · 04/10/25 · #481
LinkedIn · 04/10/25 · #481
Arxiv · 01/23/25 · #480
“Interestingly, memes created entirely by AI performed better than both human-only and human-AI collaborative memes in all areas on average. However, when looking at the top-performing memes, human-created ones were better in humor, while human-AI collaborations stood out in creativity and shareability.”
Google Docs · 03/01/25 · #480
McKinsey · 03/12/25 · #480
“Reported use of AI increased in 2024. In the latest survey, 78 percent of respondents say their organizations use AI in at least one business function, up from 72 percent in early 2024 and 55 percent a year earlier.”
Washington Post · 03/17/25 · #480
LinkedIn · 03/19/25 · #480
Quartz · 03/24/25 · #480
“But while the technology is advancing, a fundamental question remains: Should we build robots for our world, or adapt our spaces for simpler machines? The makers of these humanoid robots are pushing for the former. They argue the world is designed for human bodies, after all, with stairs, shelves at shoulder height, and important things located at eye level… They are fighting an uphill battle against the only successful robots so far, which are mostly non-humanoid robots in warehouses, where shelving systems are designed for wheeled picking robots or roped off areas that are only for robots.”
New York Times · 03/26/25 · #480
Anthropic · 03/27/25 · #480
One Useful Thing · 03/30/25 · #480
AI for Education · 04/01/25 · #480
“Miles sounds incredibly real, and quickly adapted to the direction of our conversation and often adjusted its responses to agree with us.”
New York Times · 04/04/25 · #480
McKinsey · 01/28/25 · #479
Gapminder · 02/27/25 · #479
South China Morning Post · 03/03/25 · #479
New York Times · 03/04/25 · #479
SSRN · 03/04/25 · #479
“This is the first of a series of short reports that seek to help business, education, and policy leaders understand the technical details of working with AI through rigorous testing. In this report, we demonstrate two things: There is no single standard for measuring whether a Large Language Model (LLM) passes a benchmark, and that choosing a standard has a big impact on how well the LLM does on that benchmark. The standard you choose will depend on your goals for using an LLM in a particular case. It is hard to know in advance whether a particular prompting approach will help or harm the LLM's ability to answer any particular question. Specifically, we find that sometimes being polite to the LLM helps performance, and sometimes it lowers performance. We also find that constraining the AI’s answers helps performance in some cases, though it may lower performance in other cases.”
Res Obscura · 03/05/25 · #479
“Generative AI has made it absurdly easy to generate a lot of text or images. But it hasn’t made us any better at subtracting the useful, meaningful, or simply interesting stuff from text and images — isolating the gold from the fool’s gold.”
Threads/Yann Lecun · 03/09/25 · #479
One Useful Thing · 03/11/25 · #479
“Expertise clearly still matters in a world of creating things with words. After all, you have to know what you want to create; be able to judge whether the results are good or bad; and give appropriate feedback.”
TechCrunch · 03/11/25 · #479
Uncharted Territories · 03/11/25 · #479
New York Times · 03/14/25 · #479
“I didn’t arrive at these views as a starry-eyed futurist, an investor hyping my A.I. portfolio or a guy who took too many magic mushrooms and watched “Terminator 2.” I arrived at them as a journalist who has spent a lot of time talking to the engineers building powerful A.I. systems, the investors funding it and the researchers studying its effects. And I’ve come to believe that what’s happening in A.I. right now is bigger than most people understand.”
Perplexity · 02/14/25 · #478
Marginal Revolution · 02/23/25 · #478
OpenAI · 02/02/25 · #476
“Deep research is OpenAI's next agent that can do work for you independently—you give it a prompt, and ChatGPT will find, analyze, and synthesize hundreds of online sources to create a comprehensive report at the level of a research analyst. Powered by a version of the upcoming OpenAI o3 model that’s optimized for web browsing and data analysis, it leverages reasoning to search, interpret, and analyze massive amounts of text, images, and PDFs on the internet, pivoting as needed in reaction to information it encounters.”
AI & Education · 02/10/25 · #476
“This could usher in an era of “No UI” applications, where AI agents interact with each other in the background, bypassing traditional user interfaces altogether. Welcome to the no-human web: a machine-to-machine ecosystem where AI talks to AI and humans simply orchestrate rather than directly engage. If this materializes, it could fundamentally reshape UX/UI design (not for human usability, but for AI efficiency). If AI agents do indeed handle all our browsing and subsequent interactions with information, what happens to digital literacy? Should schools be teaching students how to manage and direct AI agents? How do we develop skills to validate AI’s reasoning and choices when we’re no longer clicking, navigating, or directly engaging with content ourselves?”
Leon Furze · 02/15/25 · #476
“The length and complexity of OpenAI’s response is much greater than either Perplexity or Google, but Perplexity actually offers significantly more sources than either: 57 versus OpenAI’s 21 and Google’s 17. Obviously, quantity is not necessarily quality, but I find it interesting that there is such a marked difference. The nature of the sources is also different, with Google tending more to news stories, Perplexity with significantly more official government websites, and OpenAI drawing on blogs, news, and legal websites with a handful of government pages.”
McKinsey · 01/03/25 · #475
MIT Sloan Management Review · 01/21/25 · #475
Marc Watkins · 02/09/25 · #475
New York Times · 01/25/25 · #474
One Useful Thing · 01/26/25 · #474
Center for Curriculum Redesign · 01/01/25 · #472
“This list of 23 development vectors not only underscores the combinatorial complexity at play but also pinpoints the areas with the highest transformative potential (highlighted). Whether you’re an engineer, researcher, or enthusiast, these insights offer a roadmap to understanding GenAI's pivotal transition into its engineering phase.”
Lukas Petersson · 01/08/25 · #472
MIT Technology Review · 01/06/25 · #471
Interesting Engineering · 11/14/24 · #470
New York Times · 12/13/24 · #470
New York Times · 12/20/24 · #470
NYU · 11/22/24 · #469
“The other striking characteristic, particularly obvious if you look at the whole collection, and still more so if you have engaged with traditional formal features of poetry, either as reader or writer, is the extremely limited technical toolbox that ChatGPT uses.”
EdTechInsiders · 11/14/24 · #468
Google · 11/18/24 · #468
“We’re sharing a recap of some of the biggest scientific breakthroughs in recent years brought about by AI.”
Understanding AI · 11/05/24 · #467
One Useful Thing · 10/20/24 · #466
One Useful Thing · 10/22/24 · #466
TechCrunch · 10/03/24 · #465
The Verge · 10/04/24 · #465
Sequoia Capital · 10/09/24 · #465
Center for Curriculum Redesign · 10/01/24 · #464
“The AI community has been loose in its use of anthropocentric language, using words such as “emergence” rather than “thresholding,” “reasoning” rather than “reckoning” and “retrieval,” “understanding” rather than “pattern-matching,” and “hallucination” rather than “confabulation.” Mercifully, after a couple of years of thrashing around with definitions and counter-definitions of Intelligence, a stasis has been reached where both psychologists and AI experts seem to agree that this is a dead-end, since the concept of Intelligence is too fluid to be pinned down. But if “Intelligence” cannot be defined properly, how can “Generalized Intelligence” fare any better?”
TechCrunch · 10/03/24 · #464
AI and Education · 10/07/24 · #464
State of AI · 10/10/24 · #464
Gary Marcus · 10/11/24 · #464
The Verge · 09/25/24 · #463
Sweet GrAIpes · 10/01/24 · #463
New York Times · 09/23/24 · #462
“There are verifiable truths in the real world. A rock is a rock. Sound travels at 343 meters per second. The sun sets in the west. If A.I. systems pull information from physical reality, it can verify these truths, too. “Truth can come from the world,” Dr. Silver said. “If you can get feedback from world, you can improve and improve and improve.””
New York Times · 09/23/24 · #462
New York Times · 09/27/24 · #462
Leon Furze · 09/16/24 · #461
New York Times · 09/12/24 · #460
Anthropic · 08/27/24 · #459
One Useful Thing · 08/12/24 · #458
AI for Education · 08/15/24 · #458
McKinsey · 07/24/24 · #457
Gary Marcus · 07/28/24 · #457
One Useful Thing · 07/22/24 · #456
Facebook · 07/23/24 · #456
Gary Marcus · 07/24/24 · #456
Sweet GrAIpes · 07/24/24 · #456
New York Times · 07/19/24 · #455
New York Times · 07/19/24 · #455
Goldman Sachs · 06/25/24 · #454
Stanford · 07/01/24 · #454
Gary Marcus · 06/30/24 · #453
One Useful Thing · 07/04/24 · #453
More Than Moore · 06/13/24 · #452
Anthropic · 06/20/24 · #452
Venture Beat · 06/25/24 · #452
AI for Education · 06/28/24 · #452
Eleven Labs · 05/31/24 · #450
One Useful Thing · 06/06/24 · #450
“Every six months or so, I write a guide to doing stuff with AI. A lot has changed since the last guide, while a few important things have stayed the same. It is time for an update. This is usually a serious endeavor, but, heeding the advice of Allie Miller, I wanted to start with a different entry point into AI: fun.”
Persuasion · 06/07/24 · #450
Anthropic · 06/08/24 · #450
AI and Education · 06/14/24 · #450
Hyperallergic · 06/16/24 · #450
““I entered this actual photo into the AI category of 1839 Awards to prove that human-made content has not lost its relevance, that Mother Nature and its human interpreters can still beat the machine, and that creativity and emotion are more than just a string of digits,” Astray wrote in a blog post."
YouTube · 04/02/24 · #449
One Useful Thing · 05/12/24 · #449
Marc Watkins · 05/31/24 · #449
Brookings · 05/20/24 · #448
Anthropic · 05/21/24 · #448
EdTechInsiders · 05/21/24 · #448
Understanding AI · 05/21/24 · #448
New York Times · 05/24/24 · #448
Lance Eaton · 05/13/24 · #447
The Verge · 05/14/24 · #447
YouTube · 05/14/24 · #447
Arxiv · 01/01/24 · #446
“The aggregate forecasts give at least a 50% chance of AI systems achieving several milestones by 2028, including autonomously constructing a payment processing site from scratch, creating a song indistinguishable from a new song by a popular musician, and autonomously downloading and fine-tuning a large language model.”
New York Times · 05/08/24 · #446
New York Times · 04/29/24 · #445
McKinsey · 04/30/24 · #445
McKinsey · 04/30/24 · #445
Harvard · 05/01/24 · #445
The Verge · 04/18/24 · #444
“The Defense Advanced Research Projects Agency (DARPA) revealed that an AI-controlled jet successfully faced a human pilot during an in-air dogfight test carried out last year.”
Pragmatic Engineer · 04/21/24 · #444
The Verge · 04/23/24 · #444
The Augmented Advantage · 03/29/24 · #443
Freethink · 03/30/24 · #443
Anthropic · 04/09/24 · #443
Rolling Stone · 04/10/24 · #443
Microsoft · 04/14/24 · #443
New York Times · 04/15/24 · #443
Stanford HAI · 04/15/24 · #443
Ongoing · 04/18/24 · #443
YouTube · 03/25/24 · #442
Ars Technica · 03/27/24 · #442
One Useful Thing · 03/30/24 · #442
Gary Marcus · 03/31/24 · #442
One Useful Thing · 04/09/24 · #442
Gary Marcus · 03/10/24 · #441
Anthropic · 03/15/24 · #441
One Useful Thing · 03/04/24 · #440
New York Times · 03/11/24 · #440
New York Times · 03/17/24 · #440
One Useful Thing · 03/18/24 · #440
“For over a year, GPT-4 was the dominant AI model, clearly much smarter than any of the other LLM systems available. That situation has changed in the last month, there are now three GPT-4 class models, all powering their own chatbots: GPT-4 (accessible through ChatGPT Plus or Microsoft’s CoPilot), Anthropic’s Claude 3 Opus, and Google’s Gemini Advanced.”
State of AI · 02/11/24 · #438
“Looking back over the past five years, the report serves as a historical document. It shows both how the capabilities of the field have advanced at a dizzying pace, as well as the impact this has had on industry and politics.”
AI for Education · 02/24/24 · #438
One Useful Thing · 02/08/23 · #437
MIT Technology Review · 01/27/24 · #437
“In 1930, the prominent British economist John Maynard Keynes had warned that we were “being afflicted with a new disease” called technological unemployment. Labor-saving advances, he wrote, were “outrunning the pace at which we can find new uses for labour.” There seemed to be examples everywhere. New machinery was transforming factories and farms. Mechanical switching being adopted by the nation’s telephone network was wiping out the need for local phone operators, one of the most common jobs for young American women in the early 20th century.”
Stephen Downes · 02/06/24 · #437
Futurity · 02/08/24 · #437
“A team of researchers ran this exact experiment. They trained a multimodal AI system through the eyes and ears of a single child, using headcam video recordings from when the child was six months and through their second birthday. They examined if the AI model could learn words and concepts present in a child’s everyday experience. Their findings, reported in the journal Science, showed that the model, or neural network, could, in fact, learn a substantial number of words and concepts using limited slices of what the child experienced. That is, the video only captured about 1% of the child’s waking hours, but that was sufficient for genuine language learning.”
Byron King · 01/25/24 · #436
New York Times · 02/01/24 · #436
Chip Huyen · 10/10/23 · #435
One Useful Thing · 01/06/24 · #435
“Most likely, AI development is actually going to accelerate for a while yet before it eventually slows down due to technical or economic or legal limits. While how far AI comes this year is not yet clear, I do know that this may be the critical time to assert our agency over AI’s future. Managers, educators, and policy makers need to recognize that we are living in an AI-haunted world, and we need to both adjust to it, and shape it, in ways that increase its benefits and mitigate its harms.”
New York Times · 01/08/24 · #435
One Useful Thing · 01/16/24 · #435
Lifehacker · 01/19/24 · #435
Fast Company · 12/19/23 · #434
One Useful Thing · 12/19/23 · #434
The Verge · 12/06/23 · #433
AI for Education · 12/07/23 · #433
New York Times · 12/03/23 · #432