sohois
4d ago
Claude managed to peg me again
Finally AI is delivering what the people really want
sohois
5d ago
An LLM, strictly speaking, can't do that
That depends on how specific we want to be with "LLM". I would be surprised if ChatGPT did not have superior chess performance to someone directly calling GPT 5.4 and giving it no scaffolding beyond "We're playing chess", but few people would argue that using ChatGPT is giving an LLM extra capabilities. How much of a harness is appropriate?
If you give Claude or Codex access to a notepad skill to record any chess moves, their performances would improve. If you gave them a simple chess application with a virtual board to record moves, it would probably get even better. Would you still say that an LLM can't "strictly" augment itself in this situation?
As you yourself have pointed out, the LLM you played your simple game with independently proposed maintaining a log of game state within text, so it would be fair to say that LLMs can recognize and attempt to remedy weaknesses without human intervention, just as a normal person would do
sohois
5d ago
The chess argument is not at all a good analogy, because like a huge bunch of AI criticism, it vastly overstates normal human capacity.
What's the biggest reason an amateur teenager doesn't make impossible moves? It's because they have a chess board right in front of them. It's extremely easy to track the state of the game and position of pieces when you have the laws of physics doing it for you. How many amateurs do you think could perfectly recreate a given game state if someone came along and threw the board over?
LLMs play chess entirely through text. It's the equivalent of asking a person to play a game of correspondance chess, buth they can't recreate the game physically, they can't have any drawings of the game, all they can do is have a record of moves already made. Outside of literal chess masters, how many humans would get through such a game without making a mistake?
sohois
8d ago
AI agents are still in that awkward spot: "I spent 3 hours automating a task that would have taken me 5 minutes manually". Except previously only those skilled with computers could try this on a small range of tasks, and now anyone can do it for pretty much anything.
You absolutely could spend a week setting up Claude or Codex or whatever, putting in all the connectors and skills you need, and running a bunch of trial runs to fully automate a big chunk of your job. You'll still end up bottlenecked by stuff. Depends how much you actually enjoy working I suppose
sohois
8d ago
2 and 3: I expect the Claude Chrome add-on could handle those, depending on who each website is set-up. Two issues though: first, Claude is actually pretty slow going through Chrome, and no guarantee it wouldn't be faster to just do it yourself. Second, you're opening yourself up for all kinds of security risks at the moment.
1: As a below post says, agents can now use computers directly, but it's not great yet. Not knowing much about the situation, I expect an MCP would be enough, but it heavily depends on the software.
Tbh, the easiest solution would be to see if any of these have some kind of "Export to csv" option, then just chuck those files at Claude/ChatGPT/Whatever and tell it what you need
sohois
12d ago
I saw this Code setup in one Zvi's roundups. Might be a bit OTT for your uses though:
sohois
12d ago
Despite the name, it isn't really coding specific. I presume "Claude cowork" is largely the same as Claude code but with a name that doesn't scare the hos.
From what I can recall, Cowork was announced back when Claude Code was still only available through the terminal, and Cowork was the easy ho-accessible version. And then they make Claude Code available through the Claude app anyway, so not really clear what differences there are anymore
sohois
25d ago
Regardless, a few developments have happened recently that have motivated me enough to actually make a top-level post about this. The first being my (employer-mandated) use of Claude to generate code. "You're not using the latest model, just one more model and we'll reach AGI"-bros officially in shambles after this one
This was with the latest version of Claude Sonnet
>I totally used the latest version and it still sucked!!
>Look inside
>Not the latest version
I'm not even gonna claim Opus would make a huge difference because the differences are quite small at this point, but fuck me you would think you might have a little humility when making such an emphatic claim, just to contradict yourself within a paragraph.
sohois
27d ago
To be clear, I don't actually disagree that access to low skilled labour can suppress business investment. It's more the specific idea that this access is the biggest factor in low productivity or wage growth which I find absurd. I would be surprised if was even one of the top 5 most important factors.
sohois
27d ago
I'm not sure South Africa is the best comparison point or example for "Anglo democracies", given the unique historical factors that drive its current malaise. South Africa's democratic situation is closer to a nation like Japan, where they are essentially a one-party democracy, never deviating even in the face of catastrophe.
Plenty will argue that South Africa represents the likely future for the UK and Canada as they increasingly fracture upon ethnic lines, but there is another anglo country with massive levels of low skill ethnic minorities that is an even closer comparison - the United States. The US, not long after its inception, imported a permanent underclass that still numbers around 15% of the population, and for the past 50 years, they've had a constant influx of illegal immigration. In comparison to the rest of the anglosphere, they have a much lower % of white Europeans. Nonetheless, the US is much, much richer.
While concerns around immigration, integration, and crime are not going to be solved by money alone, South Africa's issues are clearly heavily economic in nature. The breakdown of their society is heavily influenced by the rampant corruption, the collapse of their infrastructure, and, as you say, the anaemic growth and mass joblessness. For the UK, I'd go so far as to say that the combined vote % for Reform, Restore and the Greens would be <10% if they had even kept close to the US over the past 20 years.
Both the US and Apartheid South Africa demonstrate that the economic conditions of a country are largely detached from immigration/demographics. In right-wing UK circles, I see a lot of "cope" around the plans of Reform/Restore, in which the major factor for productivity collapse is entirely low skilled immigration, and once they are kicked out companies will be forced to pay much higher wages. It's an oddly left-wing viewpoint, one in which greedy companies are keeping all the money for themselves, and you just have to force them in order to get that money to the wider public.
The reality is that the UK's pathetic productivity has been decades in the making. Clamping down on immigration levels might collapse Deliveroo and numerous Turkish barbershops, but it will not suddenly unlock hidden growth.
Most of the replies below are skeptical of saving the UK via democracy, because, I assume, they don't think that [Reform will be elected/they will try to cut immigration/they will successfully cut immigration]. I think this is the wrong viewpoint when it comes to decay or recovery. What will push UK towards South African outcomes is their complete failure to build infrastructure. It's the dead cities and towns and villages outside of London. Its the unending growth of the housing market to the exclusion of all else. It's the most expensive business energy rates in the world. And its the wages and jobs that will soon pay less than even the former communist bloc in Eastern Europe, if they exist at all.
There's not a single party that even thinks about these issues. Sure, you can find MPs and advisors that at least understand the economic woes and can propose ideas - like Danny Kruger for Reform - but even Labour and the Tories have some individuals who get it. None of them are at the centre of power, and there is such structural rot that even if they were, it would take a Herculean effort to turn things around.
So no, I don't think the UK is going to recover.
Will it decay? I'm not sure this is the truth either, more like just stagnation. There are a few bright spots for the UK: the brain drain which smashed SA is restricted for the UK. Europe is just as fucked, and so the only escape route is America. But the US has its own immigration issues, and they make it very difficult for the ~top 20-2 percentile to move there. A US that threw open the borders to white Europeans could instantly decimate most of Europe.
More than anything though, I think timescales are long enough that AI is going to render this entire conversation moot, one way or the other
sohois
29d ago
I'd say the first Opium War is easy to defend. For all that Britain wanted to force open China for trade, they had genuine grievances over treatment in Canton by the time the attack began.
But you can barely describe it as a "war". Britain peeled off a small expeditionary force from India, essentially to seek redress, probably expecting that a small show of force would be enough to bring proper negotiators to the table, instead of the belligerents that were positioned in Canton.
It turned out that the Chinese empire was in such disarray that this relatively tiny force was enough to basically sweep through the entire nation. They didn't particularly aim to cause major damage, inflict casualties, or loot the Chinese, unlike the second opium war.
I agree on the Shield. I think it came out just a bit too early to really benefit from being 'prestige TV'. Breaking Bad's popularity exploded in later seasons thanks to the growth of the internet and social media, but more than that it had the room to focus on Walt. The Shield was too long, bogged down with a lot of aimless plotlines; it felt like the production was still caught between making a serious character study, and a procedural cop show at times.
More options
Context Copy link