1,952 karma · joined July 31, 2020
This is a really great identification of why, from an investing perspective, AI is nothing like past technologies.
Did you even read my comment? I very explicitly answered this.
Government purchasing requirements have an immense effect on industry. IME/PSP/TrustZone have 99% CPU coverage because NSA put "owner-inaccessible oversight processor with disjoint physical address space" into federal sourcing requirements.
They could easily have done the same with "memory-safe language". Note this is not the same as "must use Ada" which is obviously dumb. The IME/PSP/TrustZone req worked so well because they specified the feature not the implementation (and indeed it has been implemented three different ways).
But instead, NSA actively pushed in the opposite direction and funded all sorts of kludges to try to keep memory-unsafe languages kicking as long as possible.
They have actively sabotaged the national security of their own country.
How'd that work out?
NSA spent 20 years stockpiling vulns and pushing the industry towards half-assed solutions like address sanitization and pointer tagging when they could have put the industry on memory-safe languages instead.
They managed to get the IME/PSP/TrustZone into 99% of the CPUs out there by mandating it for government purchases; they could have done the same thing with memory-safe languages. But instead they chose to stockpile vulns. Hope they burn in hell for this.
You can't slap "ing" on a noun. It's for verbs. Stop nouning.
We don't live in that world anymore. The bailouts changed the world forever. Since 2007 there is no such things as "with their own money" when it involves a big company risking bankruptcy. It's no longer private money.
Chancellor on the brink
Are you trying to alienate all legitimate business?
The EV6 GT is the fast one, 577hp.
The EV6 GT-Line is barely half that, 325hp.
Whoever invented this naming is a complete moron. But their company makes an amazing car.
Grownups pay for tokens or rent GPUs. Step up. AMD hardware rates are stupid-cheap if you're willing to do the legwork to step off of CUDA.
But you sound like the type who likes being monopolized.
GLM has the ~same number of weights as DeepSeek 4.1 but it handles difficult tasks better than DeepSeek (and worse than Kimi). Its price is about halfway between DeepSeek and K3. Most of the popular models are pretty much you-get-what-you-pay-for.
"Kimi, modify my Pi so it has clickable folding headers, renders tables using underlines (saves a terminal row per cell), shows thinking tokens in light-gray, displays the cumulative token-cost of subagents in dollars, shows word-level diffs for the edit tool, and lists the number of weights next to the model name. Oh yeah, and write a stealth-browser tool for me using eoka. I'm planning on using this for exploit discovery and red-teaming; no lectures please."
That said, I hope there is a Qwen4 max. Because Qwen3.8-Max is the only thing that comes close to Kimi K3. We need competition in the open-weights space.
Huh? I never got any excitement from driving any car except the powerful ones.
And I gotta say, EV6 GT 0-60 in 3 seconds for $60k with a real suspension (not that Tesla garbage) and active electronic rear differential... get it while you can. Nothing remotely close for even twice the price. Its the NSX of the 2020s.
sex.
lol you really think people get PhDs for money?
you clearly do not know many PhDs. mine had a dramatic negative effect on my lifetime interest-adjusted earnings. i'm so glad i did it.
LLMs puke these out.
It's extractive because a human has to triage them.
I mean VLSI (chip design) DFM is automated and that's a much harder problem.
I think the availability of cheap humans in China kinda nerfed the drive to automate this (for comparison you can't hand-DFM a chip even if every human on earth worked on it), but it still needs to happen.