tim333

21 ms·

511.
by tim333 5mo ago
Who's Gleason?
512.
by tim333 5mo ago
They know quite a lot about how neurons work to the extent they can replace bits of brains with artificial retinas or cochleas and interface with devices like neuralink. It's unlikely there is a quantum field effect of the type you men
513.
by tim333 5mo ago
Indeed, scientific type thinking like that where you ask a question that you can do experiments on to see if machines can pass the test is probably the way to progress. The philosophical waffling mostly just goes in circles. To investigate
514.
by tim333 5mo ago
>Most of the people doing this work will never encounter an important problem to solve AGI is an interesting one.
515.
by tim333 5mo ago
It probably will a bit with the Air. But I think it's good for Apple overall.
516.
by tim333 5mo ago
It's very different to an ipad and keyboard because of the software being a proper OS rather than a phone one. I was a bit torn on Neo vs Air though - the Air's a bit cooler looking.
517.
by tim333 5mo ago
There's a recognizable style to the overall piece. The subheadings for example "The summary is not the answer" and "The verification problem is real and it compounds" are typical being punchy but not really making m
518.
by tim333 5mo ago
Vaguely related, Fire Point in Ukraine just released footage of a test firing of their FP-7 rocket https://x.com/iraterekh/status/2062148065360842909 which they are hoping to make into a Patriot alternative https
519.
by tim333 5mo ago
The article: >X-Bow is an emerging player in the solid rocket motor market, using 3D printing technology to produce motors and propellants
520.
by tim333 5mo ago
That's weirdly incorrect about most stuff. I guess you are in India and reading Russian propaganda influenced sources? If Ukraine is a US proxy then how come when Trump told them to surrender and cut all funding they said yeah right an
521.
by tim333 5mo ago
There's a good wrench drop tale in https://en.wikipedia.org/wiki/Command_and_Control_(book) discussed https://news.ycombinator.com/item?id=6389633
522.
by tim333 5mo ago
In the long run stock prices are justified by sales which happen by producing things people want.
523.
by tim333 5mo ago
I have smart features on and use the web interface and am still not seeing the annoying stuff! Maybe I'm not cool enough. Or it's figured I'm not into it.
524.
by tim333 5mo ago
Also the article seems to be mixing two different things. The pricing as in >allowed to burn thousands of dollars of tokens on a $39-a-month subscription and whether the AI is worth it if you do pay what it costs. I always thought the bu
525.
by tim333 5mo ago
People worry about what will come in the future even if today's tech has limits.
526.
by tim333 5mo ago
If you're going to do a jihad it'll be easier against ChatGPT than waiting till they have Terminator 2 style things.
527.
by tim333 5mo ago
I've only heard the God narrative from skeptics as in look at those idiots thinking they are building God. I think most people who believe in AGI arriving see it more as something like a chess computer but as well as beating us at ches
528.
by tim333 5mo ago
I figure the main problem with a hard takeoff is lack of hardware capacity. The large labs are trying to improve as fast as they can but you can only do so many training runs. If you say to Claude Code try making a hundred improved versions
529.
by tim333 5mo ago
AlphaGo beat humans at go partly by studying human games but then they made MuZero that can learn games in general just through self play and became better than humans at chess, go, shogi and many others. That kind of approach my be doable
530.
by tim333 5mo ago
We don't really have superhuman intelligence yet, as in way ahead. The current stuff is ahead at some things and bad at others.
531.
by tim333 5mo ago
I think mixed economies probably work best with some central planning and some free enterprise. Like if you look at China when they sent full communist around the 1960s it was a disaster economically, but the situation from around the 1980s
532.
by tim333 5mo ago
LLMs being close to religion seems a bit of a category error like asking what shape is red and is it closer to a hexagon than green. They say it's because "LLMs rely on internal universes derived yet decoupled from reality" b
533.
by tim333 5mo ago
> And of course, The Bible has much better answers to these types of questions than science does... The Bible: >If a man has sexual relations with a man as one does with a woman, both of them have done what is detestable. They are to
534.
by tim333 5mo ago
You're not really supposed to steal oil these days and troops on the ground would be easy targets for Iranian drones. I think the military solution would to be have Ukraine help out as they have offered with land and air and marine dro
535.
by tim333 5mo ago
Probably as that gets closer their leverage will increase both due to that and stocks running down.
536.
by tim333 5mo ago
I think the idea is the salt is solid rather than in brine form. It can be sold as sea salt or whatever.
537.
by tim333 5mo ago
I figure once you get AI that can do all the stuff humans can you'll effectively get a new species whether you want it or not. Maybe the robots/AIs can go do their thing on Mars or in a Dyson sphere and Earth can be something like
538.
by tim333 5mo ago
You've got an unusual idea of what transhumanism is. I think the general understanding, as Wikipedia puts it is >Transhumanism is a philosophical movement that advocates the enhancement of the human condition by developing and makin
539.
by tim333 5mo ago
Hassabis seems decent. People have strange attitudes to tech leaders like they are all superhuman or evil or all aresholes. They are just normal people who got the top job, some good some bad, some mediocre.
540.
by tim333 5mo ago
I don't think the success is due to marketing. They've been top of the LLM Arena leaderboard for most of the year which I think is blind AB testing. Most people on HN say they are best for code. I've never seen their marketin

More ›