IdiocyInAction

I am a really heavy Claude Code user but source control is still one area where I have regular bad experiences just using Claude.

The article mentions reduced job growth in SWE due to AI but the fed actually says the opposite: https://fred.stlouisfed.org/series/IHLIDXUSTPSOFTDEVE

The people best suited for implementing and interfacing with LLMs at the moment are still SWEs and at least for the time being AI is actually probably a job creator for SWEs rather than the other way around. This might change.

And Claude has been invaluable for me to fix trade-related things at home, even complex ones. It actually outdid a locksmith!

The most resilient career is probably nursing. Medicine maybe too, not because it's not technically possible but because doctor lobbies are incredibly strong. Healthcare is the largest employer in most states now and with an aging population that's probably where much of the surplus will go and it's a profession that has really meagre productivity gains (cost disease). So nursing might be the answer.

Most of it is just blatant tax evasion.

Nurses in the US are actually very highly paid. Ask anyone how their week was. They'll all say it was crap.

Writing code can often just be the means to an end which is delivering value.

That still doesn't not make it not slavery. What if you don't agree with the social contract?

A new grad is not necessarily the same as a potential H1B hire. Tech workers are not fungible. A company might prefer to hire an Indian or Polish person who has won ICPC, has hard-to-acquire experience, etc. over a D-average new grad without internships from Georgia or something.

I think the idea of shaming people for being successful conquerers is ironically very western. Ask what the Mongols think about their conquests or the Han or what the Turks think about the Ottomans or the Russians about their history…

I don't know what you imagine perceptions of the British empire to be but most people don't think the UK sent 10000 ships to India to zerg rush the existing power structures.

He mentioned native allies at several points actually, though he doesn’t emphasise them and their purported numbers are still like 100x lower than the forces they are facing.

> And that they overemphasize planning and tactics and downplay being at the right place at the right time through no foresight of their own.

Mentioned by the author

> Another point is that the failed conquistadors are given little consideration.

Also addressed by the article.

> That book explicitly says that it is not a racist or chauvinistic history of how Europe became so successful. But critics claim that it still manages to be chauvinistic in how it wrongly claims that the conquistadors vanquished their enemies while being comically outnumbered.

I feel modern academia (and also Guns Germs and Steel FWIW) desperately try to do the opposite these days - to claim that the natives where not militarily, technologically or logistically inferior to Europeans despite getting conquered by Europeans in what look like very lopsided battles. I feel that is just as dishonest as the opposite. Nobody does this to the Mongols or Huns or whatever. Their superiority is accepted at face value.

> That's partly because inequality doesn't fit into their models.

Well Piketty being wrong doesn't fit in many people's models either. Economists routinely do talk about inequality and I think it's intellectually dishonest to paint the whole field as wrong just because some parts of it don't agree with your pet theory.

Iceland is in the EEA so EU FoM is already a thing.

AFAIK a lot of these ideas are not new (the JSON thing was done with OS models before) and OpenAI is possibly the hottest startup with the most funding this decade (maybe even past two decades?), so I think this is actually all within expectations.

> Since AI vision is not yet good enough, it’s possible that one reason the AIs do relatively poorly on this new test is because the verbal descriptions of the questions are not written as well as they could be, and further research could include multiple wordings of the questions, written by different people, to get a sense of how sensitive the AIs are to how the questions are described.

A more parsimonious theory might be that it had the other IQ test in its training set.

Still, even at that it's an impressive result; it having an actual IQ if about 100 would already be pretty substantial.

If you do a global comparison the US is incredibly richer rather than these countries being incredibly poor.

Self employment is 3 times higher in the UK than in the US actually: https://www.indexmundi.com/facts/indicators/SL.EMP.SELF.ZS/r...

The UK leads in Europe for start up founding…

Almost nobody in the UK works for the same company for 50 years and this used to be a thing in the US too.

Canada has similar laws/issues.

Social mobility is also higher in the UK than the US: https://en.wikipedia.org/wiki/Global_Social_Mobility_Index

Now I don’t trust these indices tbh, but I trust them more than a nice sounding anecdote.

There is a class divide in the UK but it's more complex/nuanced than what is described in this comment.

At least in my industry non-competes seem to be a tool to reduce worker mobility and a way for competitors to collude and keep costs down.

It's a highly-paid industry FWIW.

I think it makes sense to ban them.

Non-competes are a thing both in the UK and several EU countries and quite common in some jobs. Also quite enforceable and also not necessarily paid.

Given that women generally avoid dating lower-paid partners, there's a good chance you wouldn‘t or that those people would be outliers.

At least the version I hear most often about is a gross disparity between the genders without taking anything into account.

IIRC if you look into it it's mostly a maternity penalty anyway.

Those manufactured goods partially get enabled through these services

I get your grievances about Germany, but alimony/child support laws can turn out much worse than that in many parts of the anglosphere. Bad bureaucracy is also hardly unique to Germany.

The tax thing I do get. Especially if you are not a migrant it can be a tough ask to give away 40-50% of your money, a lot of which goes directly or indirectly to pensioners who probably hate your guts. Though again, while this is particularly bad in Germany, it's hardly unique in that regard.

Anytime Spotify makes money the music labels up their prices. Music streaming is not that great of a business model.

Yeah, I used it today to interact with some database stuff I have some passing knowledge about and it told me a bunch of wrong things. Though it also helped me solve the task; ChatGPT at least requires you to somewhat know what you are doing and to always be on your toes (I heard GPT-4 is better? Don't have access though)

Also the only fossil fuel Germany has in abundance in its own territory.

> *also to be clear I don’t mean to say achieving agi with gpt5 is a consensus belief within openai, but non zero people there believe it will get there.

Uh, that's burying the lede to say the least. "Non-zero" could mean one person.

I've copied like 3 things from SO in the last year max. Speak for yourself.

Ultra high net worth individual

I partially agree on that - for example, I find it hard to believe the LLMs will make many lawyers or doctors redundant - not necessarily because of a lack of capabilities, but due to regulatory burdens.