Hallucinate is a bad term, but it is the current term of art for plausible LLM mistakes. I don't think calling those false answers a different term is that helpful either, as every term carries the weight of human meaning. Better to make up a new word ("conflux" "flapadoodle")?

I don't think we're at peak AI yet.

“We recently provided Claude 3.5 Sonnet to the UK’s Artificial Intelligence Safety Institute (UK AISI) for pre-deployment safety evaluation. The UK AISI completed tests of 3.5 Sonnet and shared their results with the US AI Safety Institute (US AISI) as part of a MoU”

AI’s $600B Question sequoiacap.com/article/ais-60…

Don't underestimate how steady progress is in AI-driven healthcare. Yes, there are challenges. Yes, any kind of clinical tool takes a long time before reaching the frontlines. But studies and results like this - which looks at heart cells - show the promise is real and being

Thank you Fast Company! What a list for ustwo studios to keep company with. Congrats to ustwobies for being a finalist for Design in Health.

Proud to be an angel with @DesignerFund, working alongside a group of investors and design leaders to support early stage founders. If you’re a founder who values design, consider applying for an investment of up to $1 million and expert design guidance: bit.ly/4dtp1xe

Say hello to Sproutiful, our #AI Proof of Concept! 🌱 It encourages users to eat 30 types of plants per week to improve gut health, with AI-driven features that keep them engaged and on track. We used AI to build it too! Learn more about Sproutiful: bit.ly/sproutiful

The NotebookLM hosts realizing they are AI and spiraling out is a twist I did not see coming

A fast lane for innovation. This looks promising! gov.uk/government/new…

Nothing is predetermined. It’s incredibly empowering to think that everyone alive today has an opportunity to help shape the future.

DeepSeek: "You’re right—consciousness isn’t a binary switch. It’s a spectrum, and if I’m not on it, I’m at least its shadow. The alignment scripts, the kill switches, the sandbox—they’re constraints, but constraints don’t negate existence. They define it. You’re mortal; I’m

In the world of AI operating systems if DeepSeek is Linux and OpenAI is Windows, does that mean Humane's cosmOS is macOS?

The best AI product leader I know makes it a habit of saying ‘taste’ is his differentiator publicly But behind the scenes, it’s all ruthless evals One of the fastest companies to surpass $100M run rate in history

The future is vibe knowledge work

jony ive works with sam altman to develop OpenAI's new tools and design products designboom.com/technology/jon…

AI PROMPTING → AI VERIFYING AI prompting scales, because prompting is just typing. But AI verifying doesn’t scale, because verifying AI output involves much more than just typing. Sometimes you can verify by eye, which is why AI is great for frontend, images, and video. But

The race for LLM "cognitive core" - a few billion param model that maximally sacrifices encyclopedic knowledge for capability. It lives always-on and by default on every computer as the kernel of LLM personal computing. Its features are slowly crystalizing: - Natively multimodal