A bicycle for the mind

19 points by waddlesplash 15 hours ago on lobsters | 11 comments

[OP] waddlesplash | 14 hours ago

I've had the pleasure of working alongside PulkoMandy on Haiku for over a decade now, and I've learned a lot from his care and thoughtfulness in the day-to-day ins and outs of the project's workings. This post, an unpacking and a lament for what's becoming of things with LLMs, is a step up even from that.

And as someone who lives in the U.S., where car ownership is not nearly so optional (outside a select few metropolitan areas, if that) as it is in France, I think the analogy really works. I don't much like the fact that I/we're so dependent on cars to get around, but changing that at this point is nearly impossible, even if I or others wanted to. LLMs seem to be taking technology in the same direction.

sebastiancarlos | 14 hours ago

Like cars, LLMs were bound to happen eventually. We didn't get the most ethical version of them, not even close, but we can still try to improve things.

In some places there's a resurgence of clean energy and cyclepaths, so we could get a better version of LLMs one day.

5d22b | 14 hours ago

In some places there's a resurgence of clean energy and cyclepaths, so we could get a better version of LLMs one day.

If, in the metaphor, cars are LLMs, then wouldn't bicycle paths represent non-use of LLMs, not making LLMs somehow better?

sebastiancarlos | 13 hours ago

I think there are two sides here. LLMs could be better in themselves, but LLMs would also get better by there being infrastructure to not use them when we don't need to.

Halkcyon | 6 hours ago

idk, this car analogy might be apt. Selfish technology that pollutes our environment and degrades our society.

abeyer | 8 hours ago

Or you could look at it like EVs, which preserved many of the faults of cars, but still drove advances in battery tech that ebikes also end up benefiting from and wouldn't have had the market and funding to push themselves.

aarroyoc | 9 hours ago

Like cars, LLMs were bound to happen eventually.

I would like to challenge that. In order for something these things to develop, there must be a society that pushes for it. A very interesting alt-history question is why the Roman empire did not have an Industrial Revolution if they already knew basic steam engines. Or why the Spanish later, who created some steam engines similar to the later done by Watt also didn't develop it further.

Those societies didn't have the incentives to do it. They had a different set of values and priorities for them. In today's western society, productivity is one of the most important priorities and that's why LLMs are pushed for.

Sophistifunk | 6 hours ago

I think you'll find it's mostly because they had slaves.

joemi | 8 hours ago

I don't think you're challenging what the author says. You seem to be explaining why they were bound to happen eventually.

Yogthos | 10 hours ago

And the rational thing to focus on is how to develop this tech outside corporate control. I wish there were more discussion regarding training models using distributed computing or optimizing local models to be more capable while using less resources.

The oligarchs aren’t debating AI ethics, they’re investing billions to own and control this technology. Our choice is to cower in nostalgia or fight to have a stake in our future. AI won’t be stopped any more than the printing press and the internet before it. But we can have a say in how this technology is developed and who controls it.

People who weren't happy with proprietary operating systems didn't just whinge about it incessantly, they rolled up their sleeves and built their own alternatives with things like Linux and BSD. And that should be the guiding mindset today as well.

I think LLMs are a bicycle for the mind. They enable people to go further and faster than they could have before.