[simonw]

This raises an interesting question.

The amount of money that's been spent on AI related investments over the past 2-5 years really has been astonishing - like single digit percentage points of GDP astonishing.

I think it's clear to at there are productivity boosts to be had from applying this technology to fields like programming. If you completely disagree with that statement I have a hunch that nothing could convince you otherwise at this point.

But at what point could those productivity boosts offset the overall spend? (If we assume we don't get to some weird AGI that upturns all forms of economics.)

Two points of comparison. Open source has been credibly estimated to have provided over 8 trillion dollars of value to the global economy over the past few decades: https://papers.ssrn.com/sol3/papers.cfm?abstract_id=4693148 - could AI-assisted programming provide a similar boost over the next decade or so?

The buildout of railways in the 1800s took resources comparable to the AI buildout of the past few years and lost a lot of investors a lot of money, but are regarded as a huge economic boost despite those losses.

[nicoburns]

Green energy, including generation, but also storage, transmission, ev chargers, smart grid technology, etc would be the obvious thing to invest in that I would expect to have a much higher payoff.

[busterarm]

Are they really? Is this one of those "just because people say so" beliefs?

The countries adopting these the most are declining economies. It's places that are looking for something to do after there's no more oil left to drill up and export.

You know where fossil fuel use is booming? Emerging (e.g., growing) economies. Future scarcity of such resources will only make them more valuable and more profitable.

Yes, this is a dim view on the world, but until those alternatives are fundamentally more attractive than petrochemicals, these efforts will always be charity/subsidy.

If you're expecting that to be the area of strong and safe returns on investment, I've got some dodgy property to sell you.

[nicoburns]

Isn't it China that adopting renewable more rapidly than anywhere else and also has a booming economy? Although they're also investing in non-renewable energy sources.

My understanding is that "green" investment portfolios which were intended as "ethics over return on investment" have actually outperformed petrochemical stocks for years now, and it's more ideology than economics that's preventing further investment (hence why you see so much renewable energy in Texas which is famously money driven)

[MisterMower]

They are bringing neatly one new coal power plant online per day. To the extent they’re building any renewable energy capacity, it’s nuclear power, and some PV farms with their excess solar panel production that can’t be sold internationally. The Chinese are smart: they want reliable, cheap energy, and know what it takes to get it.

Renewable energy in the form of wind and solar are direct subsidies for the oil and gas industry. Wind and solar are intermittent and you have to build a proportional amount of natural gas power plants to maintain a stable grid. Every new PV farm creates demand for another natural gas power plant.

The IRA subsidies for renewables were extremely lucrative. That’s why you see renewable energy projects in Texas. Government money spends just as good as profits from oil and gas.

“Renewable” energy exists where the subsidies are, and in some rare niche cases like geothermal power in Iceland. Many of these wind farms won’t produce more energy than it took to create them.

[1oooqooq]

railways only lost investor money because everyone was investing in a national Monopoly, so the when we did get the Monopoly everyone else lost everything. sounds like a skill problem. plenty of value was created and remain in use for decades, completely different from Slop today.

[windexh8er]

Not to mention that rail only got better as more was built out. With LLMs the more you allow them to create, to scrape, and to replace deterministic platforms that can do the same thing better and faster - the further down the rabbit hole we all go.

I look around and the only people that are shilling for AI seem to be selling it. There are those that are also in a bubble and that's all they hear day in and out. We keep hearing how far the 'intelligence' of these models has come (models aren't intelligent). There are some low hanging fruit edge cases, but just again today I spent an extra hour thinking I could shortcut a PoC by having LLMs bang out the framework. I leveraged all the latest versions of Opus, Kimi, GLM and Grok. For a very specific ask (happened to be building a quick testing setup for PaddleOCR) none of them got it right. Even when asking for very specific aspects of the solution I had in mind Opus was off the rails and "optimizing" within a turn or two.

I probably ended up using about 20% of the structure it gave me - but I could have easily gone back to another project that I've done where that framework actually had more thought put into it.

I really wish the state of the art was better. I don't use LLMs for searching much as I believe it's a waste of resources. But the polarization from the spin pieces by C-levels on top of the poor performance by general models for very specific asks looks nothing like the age of rail.

Do I believe that there are good use cases for small targeted models built on rich training data? I do. But that's not the look and feel from most of what we're seeing out there today. The bulk of it is prompt engineering on top of general models. And the AI slop from the frontier players is so recognizable and overused now that I can't believe anyone still isn't looking at any of this and immediately second guessing the validity. And these are not hallucinations we're seeing because these LLMs are not intelligent. They lack cognition - they are not truly thinking or reasoning.

Again - if LLMs were capable of mass replacement of workers today OpenAI wouldn't be selling anyone a $20/month subscription, or even a $200 one. They'd be selling directly to those C-levels the utopia of white collar replacements that doesn't exist today.