The problem is the valuations assume astronomical growth... that is likely impossible for all of them to simultaneously achieve. Which means something's got to give.
4,549 karma · joined November 11, 2015
The problem is the valuations assume astronomical growth... that is likely impossible for all of them to simultaneously achieve. Which means something's got to give.
They're not and it's not clear why you seem to believe that. The immense capex for buildouts, training costs, etc. are not rolled into inference costs. Moreover, companies are already rapidly starting to re-evaluate token spend.
Honestly... not that dramatically. Each release is much more marginal. And quoted official benchmarks doesn't translate very well into the real world.
4.7 regressed hard in some ways. But a compounding factor too is that the claude code harness seems to nerf the model after a few months. Probably to reduce token use.
So far 4.8 seems less verbose but we'll see in practice what it translates into meaningfully.
To get a quality, lasting, result you're ultimately having to carefully study everything otherwise you end up quickly accumulating cognitive debt and the speedup soon shrinks as you're constantly having to revisit the initial approaches.
Increasingly, we're also seeing the moat shrink somewhat. Frontier models are converging in performance (and I bet even Mythos will get matched) and harnesses are improving too across the board (OpenCode and Codex for example).
I get why they're trying to do that (a perception of a moat bloats the IPO price) but I have little faith there's any real moat at all (especially as competitors are still flush with cash).
No matter what, common people are quickly losing agency in that discussion.
This feels a bit like a strawman. How do you assess it to be bad software without being an engineer yourself? What constitutes successful for you?
If anything, AI tools have revealed that a lot of people have hubris about building software. With non-engineers believing they're creating successful work without realizing it's a facade of a solution that's a ticking time bomb.
The reality is avoiding a ground operation was probably the wrong move at this point (ignoring the spicier broader debate of if the whole Iran campaign was the right call or not)
It's really hard to truly guarantee surface to air capabilities are gone when you're relying purely on sat images + aerial surveillance (and obviously this carries risk). Iran has fairly portable SAM systems that are public knowledge.
Avoiding working in deeply unethical areas also shields the company from legal or PR liability.
This is a REALLY hard problem that the US cannot solve alone. It would require extensive global coordination.
Not insurmountable, but this is not something you can easily roll out piecemeal. If even a single aircraft lacks the compatible equipment you're back to the existing system.
This is using hardwired weights with on-die SRAM used for K/V for example. It's WAY more power efficient and faster. The tradeoff being it's hardwired.
Still, most frontier models are "good enough" where an obscenely fast version would be a major seller.