SG team “MinD-Vis” AI model read and reconstruct images from our minds; fantastic for patients but mental privacy & “brain rights” needed to prevent decoding a person’s brain activity without consent, a new form of human rights. telegraph.co.uk/global-health/…

"AI models much smarter than us, v good at manipulating us w/o we realize it. Even if they can’t directly pull levers, they can get us to pull levers. If you can manipulate people, you can invade a building in Washington w/o ever going there yourself" mitsloan.mit.edu/ideas-made-to-…

Yes in the age of AI what makes us human is the skill of the future! Not prompt enginnering but problem formulation — the ability to identify, analyze, and delineate problems. Pb diagnosis, 5 Whys, reframing, critical thinking anyone? hbr.org/2023/06/ai-pro…

So much content at #MSBuild Great to catch OpenAI 's cofounder Andrej Karpathy State of GPT 😍🤖 youtube.com/watch?v=bZQun8…

Why AI Will Save the World by Marc Andreessen a16z.com/2023/06/06/ai-…

Microsoft’s Satya Nadella Is Betting Everything on AI The CEO can’t imagine life without artificial intelligence—even if it’s the last thing invented by humankind. wired.com/story/microsof…

Beyond the hype of #generativeai technology, I love reading about use cases and business impact per industry like this. Unsurprisingly poised to transform roles, boost perf across Customer operations, marketing & sales, software engineering and R&D mckinsey.com/capabilities/m…

Telecoms companies take first steps on emissions cuts ft.com/content/a67929… Scope 3 emissions > 98% of Tele2’s total & 90% of Telia’s. In general, biggest culprits are manufacturers of networks & networking equipment + producers of phones and mobile devices

Fav human piece "At times Robot Kyle felt like an extremely enthusiastic & productive, but rarely on-target, personal intern." newyorker.com/culture/infini… fully human-written txt luxury w AI writing on their own only historians, specialists obliged to learn reading, writing in future

The AI-Powered, Totally Autonomous Future of War Is Here wired.com/story/ai-power…

Synthetic data with loops upon loops leads to MAD (Model Autophagy Disorder), like mad cow disease. Solution? keep watermarking ON futurism.com/ai-trained-ai-…

fastcompany.com/90931084/satya…

Robots that learn as they fail could unlock a new era of AI technologyreview.com/2023/09/12/107… Lerrel Pinto says the key to building useful home robots is helping them learn from their mistakes.

scientificamerican.com/article/artifi… 2017 disco transl human lgg w/o Rosetta stone: turn semantic relations btw words into geometric ones. ML now translate by aligning shapes eg. fqcy words appear near each other to predict what comes next "ML to decode lgg that we don't know how to decode”

Tegmark wrote a landmark letter in March 2023 calling for a pause in AI dev to fully understand the dangers “[it] had more impact than I thought it would” ie political awakening on AI: US Senate hearings w tech exec, UK convening global summit on AI safety theguardian.com/technology/202…

Starting 2024 catching up on the amazing story of Microsoft and OpenAI partnership ☁️❤️🤖 newyorker.com/magazine/2023/…

AI is a people problem, not an IT problem. Why The CHRO Will Be The New Boss Of Generative AI (forbes.com) forbes.com/sites/forbesbu…

Super good Director's roadmpa for Effective AI Implementation in here issuu.com/boardprospects…

JPMorgan Chase and Capital One are winning at AI banking, expecting $2Bn ROI in the technology this year eg. Fraud detection. Quantum next frontier qz.com/ai-banking-jpm…

"The agentic era isn’t just about automating repetitive tasks. It’s about changing the way we think about work" open.substack.com/pub/hitech/p/t… capability w/o coordination not enough. Not just making AI more powerful but more usable. Smartphone needed Android, iOS, AI needs its OS.