Claude Artifacts can now be published, shared and remixed. Artifacts are awesome, and users can now share their creations to let others use or remix them.

 Why it matters  Artifacts allows anyone to create little tools within the Claude interface. Being able to use and remix others' creation is both useful and sparks ideation. Summary and demo below.

Poe introduces Previews. The new feature lets you generate and interact with web apps directly inside Poe. This is similar to Claude' Artifacts feature, but works with any of the models available on the platform, e.g. Claude 3.5 Sonnet, GPT-4o, and Gemini 1.5 Pro.

 Why it matters  We're moving from chatbots that can just give text replies, to bots that can generate small apps – think mini-games, drag-and-drop interfaces, calculators, etc. The best use cases I've seen so far are within learning (e.g. "turn this blog post into an interactive learning experience"), but we'll likely see much broader and useful applications with every upgrade.

OpenAI's 5 levels to AGI. OpenAI reportedly has a 5-level system to track the path to AGI; it goes from the currently available chatbots, to AI that can perform the same level of work as an organization by itself. We're close to the second level, "Reasoners", in which AI has the reasoning skills of a human with a PhD without access to tools. The third level is Agents (can perform actions for several days on behalf of users) and the fourth is Innovators (can invent stuff).

 Why it matters  The current level of AI already represents such a disruption in so many fields, making it all the more astonishing to consider that we might only be approaching level 2 out of 5.

Microsoft is no longer an observer on OpenAI's board. Microsoft just dropped its observer seat, while Apple has scrapped its plan to join OpenAI's board, amid antitrust scrutiny of big tech's stake in AI. OpenAI plans to maintain key partners involved through regular stakeholder meetings instead.

 Why it matters  Either, Microsoft and Apple are steering to avoid antitrust allegations (and trying to ease concerns about their influence in AI), or it could also be that OpenAI doesn't want big tech monitoring everything they do. Most of the interwebs seems to believe the former.

OpenAI partners with Thrive to launch Thrive AI Health. It’s a joint venture between the OpenAI Startup Fund and Thrive Global; the product is a hyper-personalised AI health coach app to help people make healthy behaviour changes.

 Why it matters  Many people have gotten comfortable with apps like Health on their iPhone, and this would be something similar, just with AI at the core. It's a strong use case with AI already enabling personalisation in other fields, such as tutoring. However, don't get totally blinded by these companies' focus publicly on "saving everyone's health" – remember they're also targeting a massive market opportunity.

TLDR covers the most interesting tech, science and coding news in just 5 minutes. TLDR covers the best tech, startup, and coding stories in a quick email that takes 5 minutes to read.

No politics, sports, or weather – we promise. And it's read by over 1,250,000 people. Subscribe for free (sponsored)