The state of AI

Day 167 / 366

Apple recently announced that they are bringing AI to the world. No, not Artificial Intelligence. Apple, as always, has rebranded it as Apple Intelligence. And I couldn’t be less excited.

Where are we with AI anyway? It’s been 1.5 years since ChatGPT. Has it been long enough for us to conclude that AI has hit a plateau and that this is as good as it will get?

January last year, I had thought that AI would be everywhere. Everyone will be using it in their daily lives, and you will not be able to escape it. I had thought that AI image apps would be easily available. And that everyone would be putting up AI-generated fake images of themselves.

None of that happened. Now all I use ChatGPT for is when I am too lazy to google something. And even then, half the time I have to google anyway to confirm what it tells me. Gen AI images are not realistic enough. AI influencers have not taken over the internet. All GenAI was good for, for the internet at least, are memes.

Lots of devices like the Humane AI Pin, and Rabbit R1, were exciting because they did AI stuff that your phone couldn’t do. But now our phones will be able to do all of that, so those devices are dead on arrival.

GPT-5 is nowhere to be seen. Nor is SORA.

LLMs are amazing text-complete machines. But let’s face it, they are not going to give us intelligence, let alone AGI.