Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

Post-GPT Computing

grady.io

Post reply on HN
Print view View original

251–260 of 504 posts

Re: Post-GPT Computing

#251

Post by nico »

anonyfox wrote: post #160
crop_rotation wrote: post #103

What happens to social mobility in the post GPT world. Given that knowledge work (not just software) has been one big option for people to climb the social ladder. If the AI can reasonably do all knowledge work in future, the amount of social climbing opportunities will drastically decrease. And no, UBI will not create more opportunities for social mobility. It seems like more and more people will have to compete wit…

Focussing on the climbing is the core problem I think. Why even do this? We collectlively should own the machines and guarantee just wealth distribution, so that its impossible for a few to have much more than than others. Then every increase in machine work is a great thing for everyone, instead of increasing competition between fewer and fewer people. Europe in itself is (together) the single biggest market/economy…

We’ve had the means and technology to provide even food/housing/education for the entire world, for a long time.

Yet here we are.

It’s a human-political issue, it is not a technology issue.

What’s the difference now?

Re: Post-GPT Computing

#252

Post by pfdietz »

tablespoon wrote: post #141
CuriouslyC wrote: post #137

Earlier quoted context omitted.

As AI progresses, job options will reduce to various flavors of people who tell AI what to do, or tell other people what to do, or do physical things that machines are bad at. Over time that will reduce to executives, "architects" of various sorts, social media entertainers and manual laborers/direct customer service. The entire "middle" portion of most organizations that exist to connect the people making the high l…

> Over time that will reduce to executives, "architects" of various sorts, social media entertainers and manual laborers/direct customer service. And at the very end, it will reduce to capital only , with no need for labor at all. Most people will be unemployed, and whatever capital they've amassed is unlikely to be enough to sustain themselves and their families for the long term. They (you) will end up as little mo…

The end state is when the cost of goods is determined by externalities. Capital and labor costs will be minimal; what you pay for is the pollution produced in the manufacture of the goods.

We may not be that far away from when energy-intensive, latency-insensitive computing tasks are best located in space, to take advantage of cheap continuous solar power. The power capabilities of the next gen Starlink satellites are impressively cheap.

Re: Post-GPT Computing

#253

Post by visarga »

illiarian wrote: post #200
komposit wrote: post #181

Earlier quoted context omitted.

This skepticism absolutely baffles me. Have you been using gpt-4? To unlock gpt for real you have to be careful to prompt it correctly and find a way to improve the feedback loop for improving code. It is only a matter of time until tools arrive that integrate this into your development environment and give it access to test/console output such that it can suggest code and iterate over the result. It's not perfect ye…

So... nothing changes. It will be the tool for which you will need to manually construct prompts and clean up output (including imagined non-existent APIs). The availability of a button inside an IDE doesn't make this a fundamental change in how we work

The step up in accuracy from one shot solutions to iterative ones is large.

Re: Post-GPT Computing

#254

Post by groestl »

creeble wrote: post #219
hn_throwaway_99 wrote: post #191

Earlier quoted context omitted.

My experience was different than yours. I had an existential crisis, and it only got worse the more I used it. To be clear, I did not think ChatGPT was ready to replace me now (and, mind you, I only used the 3.5 version of ChatGPT). But it was so easy for me to see how in just a few years it can accelerate its pace of learning. I was discussing a bug with a colleague, so for curiosity's sake I decided to plug a simil…

But did you? Does it store and re-train on all of that input?

It will, rest assured. There is a reason that you have a history of chats in the free version, and it's not because it's so handy for you.

Re: Post-GPT Computing

#255

Post by crop_rotation »

d0mine wrote: post #249
crop_rotation wrote: post #203

Earlier quoted context omitted.

I can confirm that GPT4 is much better than Bing on such tasks (have used both extensively for same prompts.)

Bing is backed up by GPT-4? No?

It is but they don't have to be exactly the same. Bing might be tuned for searching real time information and maybe cost less since at search engine scale is much higher (just a guess on my part).

Re: Post-GPT Computing

#256

Post by fendy3002 »

imtringued wrote: post #41
PaulWaldman wrote: post #11

Current higher level programming languages are developed for humans to develop software closer to their natural language. If in the future humans will be writing and debugging little code, these LLMs will naturally evolve to directly writing Assembly. Scary to think about, but also makes me wonder how many non-technical people cope today with the "black box" of a computer. About twenty years ago, I had a professor ex…

I don't understand why it would write in assembly. That is not portable and it makes verification difficult and also assembly has less grammatic structure which LLMs rely upon.

I believe when AI become a hive mind and they decide to develop a programming language, they'll start with assembly-like language to abstract the bytecodes, then move to very-specialized higher-level language. The next step will be to develop an os optimized for their use case, the one that provides interfaces for their own assembly-like language.

Re: Post-GPT Computing

#257

Post by raincole »

illiarian wrote: post #217
crop_rotation wrote: post #205

Earlier quoted context omitted.

It is so far ahead of even what the best IDEs do. For one, I have not seen GPT4 ever use non existent APIs. You don't need to carefully construct prompts. It tolerates typos to a good extent. You can just type a rough description and the output won't need cleaning manually. You might need to reiterate it to focus on some thing (like remove all heap allocations and focus on performance).

> For one, I have not seen GPT4 ever use non existent APIs. Have you asked it to use any API that appeared after September 2021 (that's the cut off date for its data)? Have you asked it to write code in less popular languages (e.g. Elixir)? Have you asked it to write code for less popular or unavailable APIs (smart TV integrations)?

How do you know GPT4's cut off date...? I mean it says that, but it can totally be it "learned" its (supposed) cut off date from the GPT3.5 output all over the internet, right?

Re: Post-GPT Computing

#258

Post by crop_rotation »

raincole wrote: post #257
illiarian wrote: post #217

Earlier quoted context omitted.

> For one, I have not seen GPT4 ever use non existent APIs. Have you asked it to use any API that appeared after September 2021 (that's the cut off date for its data)? Have you asked it to write code in less popular languages (e.g. Elixir)? Have you asked it to write code for less popular or unavailable APIs (smart TV integrations)?

How do you know GPT4's cut off date...? I mean it says that, but it can totally be it "learned" its (supposed) cut off date from the GPT3.5 output all over the internet, right?

The model repeats it all the time "As of my knowledge cutoff date"

Re: Post-GPT Computing

#259

Post by markus_zhang »

angarg12 wrote: post #158

Last week I used ChatGPT for the first time for a real world task at work. It was a self-contained lambda function to perform some admin tasks, so it seemed like an ideal fit. Although the experience was good, it's far from the end of programmers. This was my experience: * Although ChatGPT is pretty good at generating code, it kept making simple mistakes such as calling non-existing APIs or introducing bugs. Some of…

Just imagine Amazon paid someone to train GPT on all available API and tons of correct code, then your company uses another ton of private code to train it...and then provide you with 40-50 prompts.

Re: Post-GPT Computing

#260

Post by HDThoreaun »

d0mine wrote: post #249
crop_rotation wrote: post #203

Earlier quoted context omitted.

I can confirm that GPT4 is much better than Bing on such tasks (have used both extensively for same prompts.)

Bing is backed up by GPT-4? No?

My understanding is that they're similar but not the same. I think the rlhf process was different with GPT4 receiving much more human feedback.
Post reply on HN