I think that's true for now, but eventually there will reach a point where a model is good enough (approaching that right now with frontier models) and there will be diminishing returns. I don't need a PHD level Genius to build me an analytics dashboard for example, so why would I pay for a model with that level of intelligence when I can (eventually) self host a good enough model and run queries for electricity cost + hardware.
I think we are approaching that now, with correct expectations. With frontier large models you can often one-shot tasks with vague prompts for stuff like creating CRUD APIs and dashboards around a simple data model since it's such a solved-problem now. With something like Qwen3.6 27B or 35B-A3B and a Strix Halo level computer or a MBP with 32GB or more or RAM, you may need to be more explicit and stay involved and be a little more patient, but you can absolutely get work done with it or delegate tasks to it successfully.
My Framework Desktop does a lot of similar work as my Claude subscription at work (Cowork, chats) for 100W of power draw and a little patience waiting for a slow GPU with limited memory bandwidth to crunch the numbers. Agentic coding is obviously weaker but CRUD development and visualization dashboards are within reach, and I'm usually pleasantly surprised at its ability to self-manage devops.
I did it for a summer in University and I didn't have to work night shifts, it was amazing to be honest I would do it again if I never had to work nights.
The US had elections during their civil war in the 1800s, They had elections during WW2, major wars cannot even stop US elections legally. Doesn't mean he won't try, but it's not something he can do AFAIK.
"Not something he can do" is no longer a thing with this administration. All bets are off. Whatever logic you thought made sense, you can throw it away.
Claude 3.5 came out in June of last year, and it is imo marginally worse than the AI models currently available for coding. I do not think models are 10x better than 1 year ago, that seems extremely hyperbolic or you are working in a super niche area where that is true.
Are you using it for agentic tasks of any length? 3.5 and 4.5 are about the same for single file/single snippet tasks, but my observation has been that 4.5 can do longer, more complex tasks that were a waste of time to even try with 3.5 because it would always fail.
Yes, this is important. Gpt 5 and o3 were ~ equivalent for a one shot one file task. But 5 and codex-5 can just work for an hour in a way no model was able to before (the newer claudes can too)
I use the newer claudes and letting them work for 1 hour leads to horrible code over 50% of the time that does not work. Maybe I am not the target person for agentic tasks, all I use agents for is to do product searches for me on the internet when I have specific constraints and I don't want to waste an hour looking for something.
Your knowledge on the topic is at least six months out of date; April 2025 was a huge leap forward in usability, and recent releases in the last 30 days are at least what I would call a full generation newer technology than June of 2024. Summer 2025 was arguably the dawn of true AI assisted coding. Heck reasoning models were still bleeding edge in late December 2024. They might not be 10x better but their ability to competently use (and build their own) tools makes them almost incomparable to last year's technology.
Maybe I am just using them wrong, but I don't know how my knowledge can be out of date considering I use the tools every day and pay for Clause and Gemini. I genuinely think GPT 5 was worse than previous models for reference. They are for sure marginally better, but I don't even think 2x better let alone 10x better.
> There is a middle ground that is quite common across Europe: a private company that has the government as a shareholder.
Interestingly, Quebec in Canada has been loosely following this model recently and it's working great. Long video but interesting if you like infrastructure stuff:
TLDR: CDPQ invests on behalf of the Quebec Pension Plan, and has been building infra on behalf of the pension plan across Quebec (and Canada) to great success comparatively.
Sure but pensioners care about consistency vs. gross returns. You really don't want your pension to lose a ton of value in a downturn because people are constantly drawing from it, it's a risk off investment. Bonds are also poor investments compared to an index fund from a gross return perspective, but that's not why people/funds buy them, they buy them to lower risk.
Maybe so, but I would contend it is worth considering the broader implications of those investments and the effects that new and upgraded infrastructure could have on the greater economy.
Speaking only for myself, I would be okay with a lower return if it also means we as a society have good public transit, roads that aren't more pothole than asphalt, water that doesn't have to be boiled on occasion, reliable power, modern internet, and so on.
CDPQ has been good at making transit in Canada, but they are working on behalf of the Quebec Pension Plan so it's kind of a weird situation. I largely agree with you.
The difference in quality between model versions has slowed down imo, I know the benchmarks don't say that but as a person who uses LLMs everyday, the difference between Claude 3.5 and the cutting edge today is not very large at all, and that model came out a year ago. The jumps are getting smaller I think, unless the stuff in house is just way ahead of what is public at the moment.
And I get fast enough autcomplete results for it to be useful. I have and NVIDIA 4060 RTX in a laptop with 8 gigs of dedicated memory that I use for it. I still use claude for chat (pair programming) though, and I don't really use agents.
Oppenheimer was considered to be a left wing agitator, there are tons of very smart people who are pro labour organization which the US has historically seen as left wing agitation.
reply