Five Things AI: Sam Altman, AI Agents explained, Shenanigans, Better Code, Always Listening

Here we go AIgain!

Heya and welcome back to Five Things AI!

So right now everybody is focused on the new hardware that companies show off at CES in Las Vegas. Everybody is talking about NVIDIA and the hype train is rolling. As this is a newsletter that is not focused on news, but rather on the discussion of current trends in the AI space, let’s do exactly that: let’s look at some ideas and concepts that people take about in AI right now.

First off: what’s Sam Altman up to? Sure there’s some gossip component there when he gets asked about Musk and Trump, but aside from that it is really interesting what he has to say about AGI and OpenAI’s path towards it. You probably have seen me mentioning Agentic AI here before and I remain deeply bullish about it. I was just quoted on the popular OMR Podcast (in German) with the line: “Agentic AI will deliver what Generative AI promised.” - and that really sums up where we are right now when it comes to Agentic AI. If you want to understand what the fuzz is all about read on. Aside from the fact that I really like the word “shenanigans”, I find it fascinting that AI models are trying to come up with clever ways to trick the humans. I also find it remarkable that AI models can produce better code when you politely ask it to rewrite the code and make it better. Lastly, would you like to have an AI listen in on what you say or what people say to you all the time to then help you getting your things done during the day? I’m a bit reluctant, but the reason could also be that “Omi”means “granny” in German.

So, just dive right in, there’s a lot to unpack this week!