Weekend links: Is 2024 the year of the smart agents?

I've heard the talk about "smart, autonomous agents" on and off for years in my work as a reporter covering IT. Over the last few months more often again, and this time with a plausible road towards software that actually understand what the problem the user needs help with could be solved.

And of course, LLMs are a big part of this.

Ethan Mollick published a good article about this in December. In it, he explains how an LLM with a generalised understanding of the world could use more specialised AI-models to solve specific tasks. The smart agents are also a big part of Daniel Jefferies' predictions for AI in 2024.

More from Ethan Mollick. In Signs and Portents, he has a suggestion. Use Amara's Law together with Willam Gibson's ideas about that the future already is here, only unevenly distributed, to make some educated guesses about how AI will provide value (and challenges) going forward.

2024 is an election year in many countries around the world. If you are curious about AI in that context, Casey Newton have you covered.

Jeff Jarvis has an interesting take on AI and copyright.

If you are like me, struggling with remembering what you read and finding methods to get better at that, this is worth your time. Especially the Blank Sheet Method looks useful.