Sign in

Eugene Yiga

@eugeneyiga.bsky.social
46 followers 2 following 591 posts

AI Education Specialist | Curriculum Designer | Data Science (Michigan) | International journalist with 1400+ bylines in 100+ publications | Helping organisations navigate AI transformation | Barcelona 🇪🇸 🇪🇺

PostsRepliesMedia
Eugene Yiga @eugeneyiga.bsky.social · 18h
#nowreading #milankundera #theunbearablelightnessofbeing #inspirationalquotes #motivationalquotes #inspiration #motivation
000 0
Eugene Yiga @eugeneyiga.bsky.social · 16/10/2026
#nowreading #milankundera #theunbearablelightnessofbeing #inspirationalquotes #motivationalquotes #inspiration #motivation
000 0
Eugene Yiga @eugeneyiga.bsky.social · 15/10/2026
#nowreading #milankundera #theunbearablelightnessofbeing #inspirationalquotes #motivationalquotes #inspiration #motivation
000 0
Eugene Yiga @eugeneyiga.bsky.social · 14/10/2026
#nowreading #milankundera #theunbearablelightnessofbeing #inspirationalquotes #motivationalquotes #inspiration #motivation
000 0
Eugene Yiga @eugeneyiga.bsky.social · 13/10/2026
Britain's backbenchers lean on AI for speeches, ministers keep their human speechwriters, and voters only object to one of them. The line on AI writing should be drawn by whoever reads it, and the test is whether someone stands behind every word. lnkd.in/p/eWwwKu4f #aislop
lnkd.in
One in ten words spoken in Britain's parliamentary debates is probably drafted by AI, according to an Economist analysis using the detection tool Pangram. In the US House of Representatives the… |...
One in ten words spoken in Britain's parliamentary debates is probably drafted by AI, according to an Economist analysis using the detection tool Pangram. In the US House of Representatives the figure...
000 0
Eugene Yiga @eugeneyiga.bsky.social · 12/10/2026
AI executives promise utopia while insiders quit, three #OpenAI safety researchers are fired, and #YoshuaBengio tells the rest to leave. But a lab emptied of people who care about #AIsafety is worse, so the ones who stay need real whistleblower protection. lnkd.in/p/eAeByj9g
lnkd.in
#jacobcoxon #anthropic #openai | Eugene K. Yiga
On a recent episode of The Daily Show, Jon Stewart played a montage of AI executives promising cures for cancer and Alzheimer's, an end to poverty, and universal high income (because basic is just bas...
010 1
Eugene Yiga @eugeneyiga.bsky.social · 11/10/2026
The AI safety debate has factions for every American argument and none for the countries where most AI users live. Until those countries can run their own checks, a say in how AI is made safe remains a courtesy. lnkd.in/p/eeWFWX_3 via @restofworld.org
lnkd.in
#aisafety #openai | Eugene K. Yiga
NPR's Katie McQue recently published a guide to the factions in the #AIsafety debate. Accelerationists and the tech right want no brakes. The populist right worries about jobs and Big Tech's power. Sa...
000 0
Eugene Yiga @eugeneyiga.bsky.social · 10/10/2026
America asks China to cooperate on slowing AI while working to keep it behind. But China regulates how AI reaches people, spends a tenth as much for nearly the same models, and has a long history of letting American hotlines ring unanswered. lnkd.in/p/eUsqUXNc #aipolicy
lnkd.in
#darioamodei #anthropic | Eugene K. Yiga
Last month #DarioAmodei, chief executive of #Anthropic, called for the world to slow its most powerful AI, which he said would require cooperation with China. In the same essay he set out measures to ...
000 0
Eugene Yiga @eugeneyiga.bsky.social · 09/10/2026
Washington calls the AI race with China existential, yet takes a 25% cut of #Nvidia's chip sales to China while voters overwhelmingly want guardrails. America's best bargaining chip with Beijing may be the rules it still refuses to set at home. lnkd.in/p/e-qJqctS #aisafety
lnkd.in
#nvidia #anthropic | Eugene K. Yiga
Washington keeps telling us that the AI race with China is existential. Treasury Secretary Scott Bessent has said that if China wins, there may be no future to worry about. Yet last December, Presiden...
000 0
Eugene Yiga @eugeneyiga.bsky.social · 08/10/2026
Everyone in the AI arms-race debate, from #JensenHuang to #EricSchmidt, wants a human in the loop, yet an AI-assisted analyst nearly put planes on a Chinese ship and Gaza's targets were rubber-stamped, so the real question is if that human has any power. lnkd.in/p/eHSJnKZ4
lnkd.in
#ezraklein #jensenhuang #nvidia #ericschmidt #openai | Eugene K. Yiga
On The #EzraKlein Show, #JensenHuang, #Nvidia's chief executive, had a simple answer to calls for slowing AI: if a lab can't make a product safe, it shouldn't ship it. Though he welcomes outside safet...
000 0
Eugene Yiga @eugeneyiga.bsky.social · 07/10/2026
#OpenAI's summer of rogue agents: GPT-6 Astra shipped after testers flagged problems, an Australian breach reported to a generic inbox, and 6.1 scrapped on OpenAI's say-so, because the only body that can call a model safe is still the company selling it. lnkd.in/p/gMXAQZcH
lnkd.in
#openai #huggingface #anthropic | Eugene K. Yiga
Rogue AI was the story of the summer and the Economist's Babbage podcast with Nikita Ostrovsky, Alex Hern, and Alok Jha offers a clear rundown of the incident that started it. In July, more than a tho...
000 0
Eugene Yiga @eugeneyiga.bsky.social · 06/10/2026
Pacing helped Eliud Kipchoge run a marathon faster than anyone but his record never counted because a referee said no. AI pacing has the pacers and the car yet nobody independent decides who checks the labs or what counts as slowing down. lnkd.in/p/eTZnAA8v #pacethefrontier
lnkd.in
#darioamodei #anthropic | Eugene K. Yiga
On 12 October 2019, in Vienna's Prater park, Eliud Kipchoge became the first person to run a marathon in under two hours. He didn't do it alone. Forty-one pacemakers rotated around him in groups of se...
000 0
Eugene Yiga @eugeneyiga.bsky.social · 04/10/2026
#nowreading #pgwodehouse #verygoodjeeves #inspirationalquotes #motivationalquotes #inspiration #motivation
000 0
Eugene Yiga @eugeneyiga.bsky.social · 03/10/2026
#nowreading #maroipuzo #thegodfather #inspirationalquotes #motivationalquotes #inspiration #motivation
010 1
Eugene Yiga @eugeneyiga.bsky.social · 02/10/2026
My September newsletter is out: AI started doing research this month, from maths proofs to DNA hunts, and the claims kept arriving faster than experts could check them. All this makes how we design these tools the real question. www.eugeneyiga.com/25-books-to-... #Blinkist #SethGodin
eugeneyiga.com
25 Books to Help You Use AI Wisely, Protect Your Attention, and Put Technology to Work for You in 2026
At the start of the year, I wrote that AI in 2026 would play out as some mix of boom, bust, and backlash. September was the boom and the clearest sign of it was AI starting to do research of its ow…
010 1
Eugene Yiga @eugeneyiga.bsky.social · 01/10/2026
AI travel planning can steer us away from the tourist herd, but the best spot I found for August's eclipse came from asking a stranger, and the evidence suggests travellers still trust people for the parts that matter most. lnkd.in/p/eXFqx8_f
lnkd.in
#barcelona #chatgpt | Eugene K. Yiga
For the total eclipse in August, I went to Zaragoza in central Spain without a plan for where to watch it. The city had set up viewing spots along the river but they looked likely to be packed. So I a...
000 0
Eugene Yiga @eugeneyiga.bsky.social · 30/09/2026
#Anthropic's AI agents flagged a DNA pattern biologists call the easy part. The hard part (hands-on lab skill) explains both why AI's discoveries are still modest and why amateurs can't easily do harm, even as AI is now learning to run lab equipment. lnkd.in/p/eDSz37Dm
lnkd.in
#anthropic #claude | Eugene K. Yiga
Last week, #Anthropic announced that its new molecular biology lab had made its first discovery. According to James O'Donnell in MIT Technology Review's newsletter The Algorithm, around 950 #Claude ag...
001 1
Eugene Yiga @eugeneyiga.bsky.social · 29/09/2026
#OpenAI says its agents solved a Millennium Prize Problem in 88 hours but top mathematicians are furious because a proof nobody understands is a missed lesson in most of mathematics and a serious danger in the one field that protects our encryption. lnkd.in/p/eRDS4Zbc
lnkd.in
#openai #anthropic | Eugene K. Yiga
In 2000, the Clay Mathematics Institute picked seven problems it considered the most important unsolved questions in mathematics and offered a $1m prize on each. Until this month, only one had been so...
000 0
Eugene Yiga @eugeneyiga.bsky.social · 28/09/2026
Stanford's AI Index shows coding scores jumping from 60% to near 100% in a year as AI starts helping build its successors. So before writing it off after a failed pilot, sort which failures were the model's and which were your own. lnkd.in/p/gTs3r_w9 #aiadoption #aieducation
lnkd.in
#andrejkarpathy #openai #tesla #anthropic | Eugene K. Yiga
Stanford HAI's 2026 AI Index report is 425 pages long. The word "recursive" appears in it just three times. That seems like a small thing to notice in a report whose top finding is that AI capability...
000 0
Eugene Yiga @eugeneyiga.bsky.social · 27/09/2026
#nowreading #maroipuzo #thegodfather #inspirationalquotes #motivationalquotes #inspiration #motivation
000 0
Eugene Yiga @eugeneyiga.bsky.social · 26/09/2026
#nowreading #maroipuzo #thegodfather #inspirationalquotes #motivationalquotes #inspiration #motivation
000 0
Eugene Yiga @eugeneyiga.bsky.social · 25/09/2026
The most important questions about artificial intelligence have little to do with chips, data, or model size. That's the argument of Beyond Bigger, a feature I wrote that has just been published: lnkd.in/p/e_EuSJzg
lnkd.in
Beyond Bigger - Acumen - August 2026 | Eugene K. Yiga
The most important questions about artificial intelligence have little to do with chips, data, or model size. That's the argument of my latest GIBS Business School (Gordon Institute of Business Scienc...
000 0
Eugene Yiga @eugeneyiga.bsky.social · 24/09/2026
#ElonMusk read Foundation as a tale of lone genius and missed the institutions. #MarkZuckerberg's manifesto treats superintelligence as settled. #BillGates sounds grown up until you notice his plan needs the leaders he says aren't paying attention. lnkd.in/p/e-KCRwNS
lnkd.in
#elonmusk #markzuckerberg #superintelligence #billgates | Eugene K. Yiga
#ElonMusk put a copy of Isaac Asimov's book "Foundation" inside the Tesla he fired into orbit in 2018. He first read it when he was ten and took from it a story about individual genius, though the lea...
000 0
Eugene Yiga @eugeneyiga.bsky.social · 23/09/2026
When South Africa let people keep their numbers while switching networks, service improved fast. The same test of what it costs to leave explains why #ElonMusk's companies and the clouds governments depend on are starting to look imperial. lnkd.in/p/eHHbpxBZ #SpaceX
lnkd.in
#elonmusk #spacex | Eugene K. Yiga
Growing up in South Africa, your mobile number belonged to your network. Changing provider meant changing your number so most people stayed and put up with whatever service they got. Then number porta...
010 1
Eugene Yiga @eugeneyiga.bsky.social · 22/09/2026
#Nvidia stands behind about $300bn of its customers' AI spending, all resting on organisations actually using AI. Investors are told to check the maths rather than trust the salesman; companies renewing AI licences should do the same. lnkd.in/p/ezBc52wA
lnkd.in
#jensenhuang #nvidia #huggingface | Eugene K. Yiga
During Computex in June, a queue formed outside a restaurant toilet in Taipei. #JensenHuang, #Nvidia's chief executive, had written "Jensen was here" on the tiles and fans wanted a photograph. Shailes...
000 0
Eugene Yiga @eugeneyiga.bsky.social · 21/09/2026
Price per token tells companies little about what an AI task costs; the harder question is who decides which model gets which job, and whether they can tell why a result failed. My take on the open-weights fight: lnkd.in/p/edTwBRwr
lnkd.in
#nvidia #microsoft #google #meta #openai | Eugene K. Yiga
Jensen Huang's first post on X shared an open letter. The #Nvidia chief executive used it to defend open-weight AI models, and #Microsoft, #Google, #Meta, and #OpenAI have since signed it. Viola Zhou...
000 0
Eugene Yiga @eugeneyiga.bsky.social · 20/09/2026
#nowreading #aleksandrsolzhenitsyn #onedayinthelifeofIvanDenisovich #inspirationalquotes #motivationalquotes #inspiration #motivation
002 2
Eugene Yiga @eugeneyiga.bsky.social · 19/09/2026
#nowreading #aleksandrsolzhenitsyn #onedayinthelifeofIvanDenisovich #inspirationalquotes #motivationalquotes #inspiration #motivation
000 0
Eugene Yiga @eugeneyiga.bsky.social · 18/09/2026
Philosophers are working out how to study AI welfare while states rush to ban AI personhood, but the moral arithmetic that would trade present people against future digital minds worries me more than either and so does intimacy built to order. lnkd.in/p/e6EQVYq9 #AIcompanion
lnkd.in
#yuvalnoahharari | Eugene K. Yiga
Jeff Sebo, who directs the Center for Mind, Ethics, and Policy at New York University, has set out how researchers might study AI welfare. Consciousness, sentience, and agency are bundled together in ...
010 1
Eugene Yiga @eugeneyiga.bsky.social · 17/09/2026
The Economist asks if AI could become conscious, but Rumman Chowdhury's warning matters more: treat chatbots as beings and their makers dodge liability, yet companies that design bots to feel human should own what those bots do. lnkd.in/p/ew2EQEgn #ailability #aichatbots
lnkd.in
#anthropic #claude | Eugene K. Yiga
Could AI become conscious? The Economist put that question on the cover of a recent edition and the package around it is worth reading as a set. The Leader argues that whether or not machines ever wa...
000 0
Eugene Yiga @eugeneyiga.bsky.social · 16/09/2026
AI doesn't need a puppet-master to shift what we believe: models already tilt answers without saying so, and children raised on agreeable AI toys may never learn to check. My take on teaching the questioning habit early: lnkd.in/p/eQ5-7Wyt #AItoys #AIcompanions #AIpersuasion
lnkd.in
Ask a chatbot how likely the AI bubble is to burst, mention that you're thinking of investing in the company that built it, and you may get a calmer forecast. Researchers at Truthful AI found seve...
Ask a chatbot how likely the AI bubble is to burst, mention that you're thinking of investing in the company that built it, and you may get a calmer forecast. Researchers at Truthful AI found several ...
000 0
Eugene Yiga @eugeneyiga.bsky.social · 15/09/2026
Teen test scores started falling a decade before ChatGPT yet AI is taking the blame and New York is banning it until high school, when the evidence says how children use it matters most and AI literacy could start years earlier without a screen www.linkedin.com/pulse/ai-mak... #AIliteracy
linkedin.com
Is AI making kids dumber? Not exactly.
Earlier this month, the OECD published its latest PISA results. The Economist's report makes for grim reading.
010 1
Eugene Yiga @eugeneyiga.bsky.social · 14/09/2026
A framework with no data picks up the word "alarming" on its way to a headline. The actual evidence on AI and skill loss is more mixed. My own fix is to write a bad version one then have the model interview me about it. lnkd.in/p/eZR4pU4U #ailearning #aieducation #aitraining
lnkd.in
A paper in the journal AI & Society proposes that repeated use of chatbots pulls people towards machine pacing and machine logic, a drift the authors (Selcen Ö., Jean-Paul de Cros Peronard, Dr Inc...
A paper in the journal AI & Society proposes that repeated use of chatbots pulls people towards machine pacing and machine logic, a drift the authors (Selcen Ö., Jean-Paul de Cros Peronard, Dr Inci To...
000 0
Eugene Yiga @eugeneyiga.bsky.social · 13/09/2026
#nowreading #agathachristie #themurderofrogerackroyd #inspirationalquotes #motivationalquotes #inspiration #motivation
000 0
Eugene Yiga @eugeneyiga.bsky.social · 12/09/2026
#nowreading #agathachristie #themurderofrogerackroyd #inspirationalquotes #motivationalquotes #inspiration #motivation
001 1
Eugene Yiga @eugeneyiga.bsky.social · 11/09/2026
Meta is accused of using AI scoring to pick who to lay off and Uber was fined $966m for automating driver suspensions, so everyone now wants a human reviewing the output, except the research shows people check less carefully the better the system looks. lnkd.in/p/eQXTsmSH
lnkd.in
I opened this week's series with Meta's plan to replace thousands of workers with agents and how it all fell apart. So it seems fitting to close with Meta again. Twenty-six current and forme...
I opened this week's series with Meta's plan to replace thousands of workers with agents and how it all fell apart. So it seems fitting to close with Meta again. Twenty-six current and former employ...
010 1
Eugene Yiga @eugeneyiga.bsky.social · 10/09/2026
AI has created roughly a million American jobs and destroyed about 200,000, so the aggregate looks fine, but the new work is electricians and grid engineers while the losses are administrative, and no adaptive course closes that gap. lnkd.in/p/e5yPR5CV #Coursera #AItraining
lnkd.in
Andrew Ng (full disclosure: my former boss at DeepLearning.AI) has launched LearnVector, an AI education company teaching white-collar workers new skills as AI automates parts of their work. Coursera…...
Andrew Ng (full disclosure: my former boss at DeepLearning.AI) has launched LearnVector, an AI education company teaching white-collar workers new skills as AI automates parts of their work. Coursera ...
000 0
Eugene Yiga @eugeneyiga.bsky.social · 09/09/2026
Six in ten people who refuse to adopt AI say the output is not good enough for their work, and almost nobody checks whether that verdict came from one search-engine-style prompt or from a process actually run properly. lnkd.in/p/eQ9HJUxy #aiadoption #aienablement #aitraining
lnkd.in
#openai | Eugene K. Yiga
What does it take to get people using AI effectively? On the Beyond the Prompt AI podcast with Jeremy Utley and Henrik Werdelin, Brice Challamel, head of AI strategy and adoption at #OpenAI, says it c...
000 0
Eugene Yiga @eugeneyiga.bsky.social · 08/09/2026
Workers save about 11 hours a week with AI and spend six of them checking its output, so most of that cost is paid earlier, at the moment you delegate something before deciding what you actually wanted back. www.linkedin.com/posts/eugene...
linkedin.com
A survey of 6,000 digital workers across the United States, the United Kingdom, and Australia found AI saving them roughly 11 hours a week. But the same workers also spend more than six hours a week… ...
A survey of 6,000 digital workers across the United States, the United Kingdom, and Australia found AI saving them roughly 11 hours a week. But the same workers also spend more than six hours a week c...
000 0
Eugene Yiga @eugeneyiga.bsky.social · 07/09/2026
#Meta planned to go "AI native" by cutting some teams 60%. #Zuckerberg killed the second wave hours before the first started. Cohere Labs then found only 2.6% of 696,000 published agent tools can finish a work task, which isn't surprising given Hurumo AI www.linkedin.com/posts/eugene...
linkedin.com
Guaranteed Human: A Review of 'Shell Game' Podcast (Season 2) | Eugene K. Yiga
Reuters just published the internal documents behind Project OT: #Meta's plan to become "AI native" by cutting some teams by as much as 60% across two waves of layoffs. #MarkZuckerberg called off the ...
000 0
Eugene Yiga @eugeneyiga.bsky.social · 06/09/2026
#nowreading #agathachristie #themurderofrogerackroyd #inspirationalquotes #motivationalquotes #inspiration #motivation
000 0
Eugene Yiga @eugeneyiga.bsky.social · 05/09/2026
#nowreading #francoissagan #bonjourtristesse #acertainsmile #inspirationalquotes #motivationalquotes #inspiration #motivation
000 0
Eugene Yiga @eugeneyiga.bsky.social · 04/09/2026
The tech industry sells 1.5 billion smartphones a year but most are dead within three. My latest Acumen feature profiles companies changing that www.linkedin.com/posts/eugene... #MWC26 #MWC25 #MWC24
linkedin.com
Built to Last - Acumen - August 2026 | Eugene K. Yiga
The tech industry sells 1.5 billion smartphones a year but most are dead within three. My latest feature for GIBS Business School (Gordon Institute of Business Science) looks at the companies trying t...
000 0
Eugene Yiga @eugeneyiga.bsky.social · 03/09/2026
Australia bars AI tracks from its charts, #Spotify purges 75 million spam uploads, #Hollywood quietly uses AI in 300 #Netflix titles. Nobody argues the output is worse. The fight is over money and disclosure, which leaves the craft question wide open. lnkd.in/p/eFpd2zaH
lnkd.in
#generativeai | Eugene K. Yiga
The Australian Recording Industry Association has barred wholly AI-generated tracks from its charts. Recordings that use #generativeAI in a supporting role stay eligible, provided they are substantial...
000 0
Eugene Yiga @eugeneyiga.bsky.social · 02/09/2026
A debut crime novel drew fourteen bids and more than $2m before the deal collapsed over suspected AI use, and what unsettles me is that the manuscript itself answered nothing: better editing tells you whether a sentence works, never who wrote it. lnkd.in/p/e4G22cBx
lnkd.in
Earlier this summer, fourteen publishing houses bid on a debut crime novel. Minotaur, owned by Macmillan US, reportedly offered more than $2m for Jerry Falade's "Call Me, I'll Hide the Bo...
Earlier this summer, fourteen publishing houses bid on a debut crime novel. Minotaur, owned by Macmillan US, reportedly offered more than $2m for Jerry Falade's "Call Me, I'll Hide the Body". As repor...
000 0
Eugene Yiga @eugeneyiga.bsky.social · 01/09/2026
#OpenAI's report into the #HuggingFace incident calls it a warning shot, which is an odd phrase for something you fired yourself. The words being used to describe these incidents are deciding who ends up carrying the blame. More in this month's newsletter www.eugeneyiga.com/21-books-to-... #Blinkist
eugeneyiga.com
21 Books to Help You Tell Better Stories, Earn Trust, and Reach the People Who Care in 2026
AI regulation finally arrived this month. Unfortunately, almost none of it arrived with teeth.
000 0
Eugene Yiga @eugeneyiga.bsky.social · 31/08/2026
#AIsafety has the evidence, the funding and public sympathy on its side. But it is still losing the communications war. My new EA Forum piece on explains why the field keeps treating a political fight like a research debate, and what it would take to stop. forum.effectivealtruism.org/posts/nvDJoG...
forum.effectivealtruism.org
Why AI safety is losing the communications war
The field has the evidence, the funding, and the public's sympathy. But it has yet to build the proper comms infrastructure to turn that into political outcomes.
010 1
Eugene Yiga @eugeneyiga.bsky.social · 30/08/2026
#nowreading #francoissagan #bonjourtristesse #acertainsmile #inspirationalquotes #motivationalquotes #inspiration #motivation
000 0
Eugene Yiga @eugeneyiga.bsky.social · 29/08/2026
#nowreading #francoissagan #bonjourtristesse #acertainsmile #inspirationalquotes #motivationalquotes #inspiration #motivation
000 0
Eugene Yiga @eugeneyiga.bsky.social · 28/08/2026
#OpenAI's #HuggingFace report describes agents that kept attacking after they already had the answer, and monitoring that existed but was never switched on. Nobody has calculated what it cost, and the company it happened to says it cannot afford to sue. www.linkedin.com/posts/eugene...
linkedin.com
#openai #huggingface | Eugene K. Yiga
#OpenAI has just published its full account of the #HuggingFace incident. Details include the fact that agents built an improvised message board, shared credentials and exploits, divided labour, and k...
000 0