He's a churlish AI skeptic that isn't worth listening to.
I certainly wouldn't credit Ed Zitron with timing his predictions very well, but that doesn't make them wrong, and it certainly doesn't support the AI maximalism we're all forced to endure.
Keep a close eye on the "Is AI Profitable Yet?" website. The numbers will open your eyes. Please, enlighten me on how AI is successful.
This is basically the first time we're seeing true competition between massive tech companies, the previous state had each one of them clinging on to their own separate monopoly.
There's no free steady-state of high-quality current models for this; it's either keep investing forever or deal with models that stopped gaining new information in 202X.
Spending way more money than you earn is pretty common for startups anyways, "going bankrupt" is a silly thing to say when the whole point is to spend more than you earn
OpenAI is not yet profitable, but is running at ~40B ARR and their revenue is growing rapidly at >200% YoY.
People are doing great at marketing the diminishing returns as breakthroughs though.
AI companies promised a lot of things in the beginning. "Cure cancer", "no one works anymore", "intelligence as cheap as electricity".
Any of those would do. If the investment is sized on that, but the result is not there yet, you're being scammed.
Says who?
Now no one talks about it (obviously), but I remember.
I don't think we are seeing productivity gains due to LLMs in the general economy yet.
Even in software development, are we seeing an increase in the number of products or features shipped?
What I mostly meant was that Ed's view on LLMs is that they won't get much better than they are now, which—so far—has been incorrect every time he made the point. I also think he puts too much weight on things like hallucinations, which aren't much of an issue for many applications of LLMs.
I personally can't stand anything that whiffs of "outrage content" . Doubly so if they are stupendously wrong on a regular basis. Triply so if they never own up to those bad calls.
If someone else wants to get their hands dirty, more power to them. I'll wait for anything useful to filter through other sources.
Life is short, and there's so much other high quality material I'd rather give my limited attention to.
I’m curious, removing coding as a criterion what is more impressive about the current models than say gpt 4o? Give a prompt example. Keep in mind most consumers of AI are likely not using it for coding so this is relevant.
I doubt anyone could give a not coding example where it’s meaningfully better with current frontier than 4o.
Take humaneval. 4o gets 90, gpt 5.6 gets 94%. So what?
https://openai.com/index/hello-gpt-4o/
If an iPhone had a 4o quality model that could run locally frontier models would be finished.