☆ Yσɠƚԋσʂ ☆

  • 15.2K Posts
  • 14.4K Comments
Joined 6 年前
Cake day: 2020年1月18日
  • ☆ Yσɠƚԋσʂ ☆@lemmy.mlOPtoTechnology@lemmy.ml•Running local models is good now
    2·
    3 小时前

    Having done development for over two decades now, I’m really not learning anything useful when I make yet another CRUD end point on a server, or a new widget. The reality is that most coding tasks are highly repetitive and we’re just writing the same boiler plate in slightly different contexts. Being able to offload boring and repetitive tasks to a machine is what automation is for.

    I’d rather spend my brainpower on things I find interesting like the overall architecture and the problem being solved while leaving writing implementation details to the LLM. It’s not like you stop solving problems when you use an LLM for coding, you’re just focusing on different things at that point.

    It’s also worth noting that this argument isn’t new. I’m old enough to remember how writing assembly by hand was what real coders did or how using GC was cheating because you shouldn’t offload memory management to the computer. In each case it turned out that using better tools let us build more interesting things in the end and freed up human thinking from boring and repetitive work.

  • ☆ Yσɠƚԋσʂ ☆@lemmy.mlOPtoWorld News@lemmy.ml•‘Netanyahu’s life project failed with US-Iran deal’
    1·
    3 小时前

    The US wouldn’t be capitulating right now if they were in a position to continue the war, and Israel can’t do shit on its won. They will absolutely try to keep the war going, but they’re running up against material limits here. I also doubt Trump himself really matters all that much here, there’s a huge amount of political capture by Israel in the US, and this lobby is going to push for war no matter what. But US economy and military industry isn’t capable of continuing it.

  • ☆ Yσɠƚԋσʂ ☆@lemmy.mlOPtoWorld News@lemmy.ml•Russia Ships Oil at Near-Record Pace as Kyiv Pummels Refineries
    52·
    17 小时前

    Pinprick attacks on refineries don’t actually have a big impact. They get a lot of media, but they don’t actually cause major disruptions for more than a few days. If you look at the size of oil refineries you’ll quickly realize that a single drone isn’t going to do much to them. This sort of propaganda is aimed at people who have no clue how this stuff actually works, and just look at pictures of smoke and think Russian oil production has completely stalled while in reality it’s largely unaffected.

  • ☆ Yσɠƚԋσʂ ☆@lemmy.mlOPtoTechnology@lemmy.ml•Running local models is good now
    2·
    22 小时前

    I don’t mean you turn the program itself into a genetic algorithm. I’m saying that the agentic loop for producing code acts as one. The code itself is just regular code. And the loop isn’t really any more inefficient than what you do as a developer. It almost never happens that you write perfect code on a first try in practice. You’ll write some code, run your tests, look how it did, and iterate. That’s precisely the same process the agent follows.

    The difference from a typical genetic algorithm is that the LLM is not just randomly generating text that eventually fits into the shape you specified. It’s generating code that’s already close to what’s intended most of the time, and it just needs a bit of massaging to get completely right. That’s the feedback loop here.

  • ☆ Yσɠƚԋσʂ ☆@lemmy.mlOPtoTechnology@lemmy.ml•Running local models is good now
    4·
    23 小时前

    I find I kind of look at the whole agentic harness setup as a genetic algorithm. Your tests and specs are the fitness function for the program you’re evolving, and the LLM is the mutator. At each step it generates some output, it gets tested against the fitness function, the LLM gets feedback and iterates on it. Eventually something working falls out in the end. The better you can define the selection criteria the more you box the agent in the better results you get.

    The trick I can recommend for getting the model to code is to ask it to come up with a phased plan composed of focused features, and then to build each feature on its own branch. That way you have a clear unit of work that does a specific thing which makes it much easier to review the code. Can also recommend tools like https://github.com/Fission-AI/OpenSpec for making specs to box the model in when it works.

  • ☆ Yσɠƚԋσʂ ☆@lemmy.mlOPtoTechnology@lemmy.ml•Running local models is good now
    10·
    1 天前

    You can run the Gemma 4 and Qwen3.5 MoE models with as little as 12 GB of VRAM at 30-40 tps (Q4/Q5), and they both blow GPT-4o and DeepSeek R1 out of the water. But 64gb RAM is also not really out of scale with the cost of a shop tool in other trades. If you’re a professional that’s confident in a positive return on the investment, or just a hobbyist with the luxury budget for a “shop” that cost is well within consumer market. That’s not everybody, of course, but it’s not some inconceivable fantasy.

    The key point is that local models continue to get more efficient and usable. You need high end consumer grade hardware today, but given how fast improvements are happening, it’s entirely likely that you’ll be able to get the same capability on even smaller hardware in a few months.

  • ☆ Yσɠƚԋσʂ ☆@lemmy.mlOPtoSocialism@lemmy.ml•Everything belongs to the workers
    1·
    19 小时前

    I think Stalin was largely correct in what he did, the problem was that he left a system which failed to ensure strong leadership going forward. A stable social system can’t depend on a single strong willed individual being in charge and making the right calls. Continuity of competent governance, especially in time of plenty is the hardest problem to solve in my opinion.

    And completely agree, China quietly outplayed the west. A lot of it was inherent in western hubris too. They really thought that theirs was the only way to develop, and they figured that China would have to become like them eventually and they’d fold it in. But it didn’t work out that way. Turns out people with 3000 years of continuous civilization under their belt know a thing or two of their won. Also, don’t know if you saw, but American media has now realized DPRK is doing rather well. https://archive.ph/b9zrS

    The west really is starting to look like the final days of the Roman empire now. I expect we’ll start seeing provinces getting cut loose next and imploding economically. The UK looks like it might be the first to pop.

    oh and just ran across this https://global.chinadaily.com.cn/a/202605/19/WS6a0c0718a310d6866eb4976d.html

  • ☆ Yσɠƚԋσʂ ☆@lemmy.mlOPtoSocialism@lemmy.ml•Everything belongs to the workers
    1·
    1 天前

    I don’t think the crisis has been averted. It’s going to take a long time before energy prices get back to normal because restarting production can’t happen overnight. Just clearing the backlog of tankers in the gulf is going to take over a year. I also don’t see Israel stopping attacking Lebanon which means the fighting is likely to restart soon.

    Trump wants to get out desperately, but he has no way out because Israel won’t play along. From Russian perspective it makes sense to play along though because it drives Europeans up the wall. And I don’t see what leverage he has left either.

    And completely agree that strikes just serve to remind people in Russia why the war is necessary. The overall situation on the front won’t change, but it will help with firming up public support to remove the threat.

    It does look like Russia is ramping up deep strikes on infrastructure especially now that the US ran out of patriots during their Iran fiasco. I think this will be significant over time, and affect logistics going forward which will accelerate the events on the front.

    I saw a video just yesterday of some kid beating up TCK cause they took his dad. Yes, public is definitely starting to turn on them.

    And American style propaganda does in fact have its origins with Goebbels, I might’ve sent this before. It explains everything very clearly. https://royallib.com/read/artemov_vladimir/psihologicheskaya_voyna_v_strategii_imperializma.html#0

Moderates