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

GPT is all you need for the back end

github.com

Post reply on HN
Print view View original

271–278 of 278 posts

Re: GPT is all you need for the back end

#271

Post by bufferoverflow »

goatlover wrote: post #92
bufferoverflow wrote: post #68

Earlier quoted context omitted.

You are factually wrong. The general case is being solved. Self-driving systems are objectively better every year, and will eventually reach human level safety. You can literally watch cars self-driving in all kinds of places and conditions. Yes, they make mistakes. So do humans.

So when will these self-driving cars be ready for mass consumption? Because they're starting to sound like flying cars at this point. Yes, technically doable and there are always some working prototypes, but no real market, and not seen as practical.

We have flying cars.

Re: GPT is all you need for the back end

#272

Post by candiodari »

krainboltgreene wrote: post #163
oceanplexian wrote: post #150

Earlier quoted context omitted.

I keep hearing this assertion, that GPT can be wrong, therefore it’s an unworkable technology. But it’s a bad comparison. LLMs aren’t trying to be computationally correct like a calculator or something, the value is in their ability to semantically process a question. The other issue is assuming that the existing way of doing things is always correct. Engineers frequently get things wrong. If an AI model can complete…

> If an AI model can complete a task with 95% correctness but let’s say a Jr. Engineer can compete the same task with 85% correctness then it makes sense to use the model instead. I’m not sure why folks can’t see the obvious conclusion of where this is heading. Because this is incredibly shortsighted and also fundamentally misunderstands the return data of an LLM.

Plus you need to see ChatGPT fail. When it fails, it will keep actively trying to persuade you it's right rather than fix problems.

It does this through 2 deceptive techniques, which no doubt work on many people:

1) it suggests it was right all along, very politely pressuring you to accept it. It does not really make arguments for why it's right. It just pushes you to accept it's output. (if a human does this, I would argue this is a human trying to hide the fact they made a mistake, or perhaps hiding they don't know. Not an honest mistake. Either way, VERY bad sign)

2) (and/or) it will suggest and make changes that don't address the concern. In a way this is the same as 1), but ...

Using ChatGPT for anything remotely important runs a very, very big risk of causing disasters imho. For a template, basic inspiration, perhaps ... but ...

Let's put it this way. If you gave control of a nuclear plant to ChatGPT, it would make you feel good about this, then melt the plant down.

ChatGPT is incredibly impressive. But it's a con man. Hell, it's a better con artist than a lot of human con men, but it's fundamentally trying to convince you, not caring if it's right or wrong. It's a "troll". An incredibly good troll. But it's not trying to solve problems. Extremely impressive achievement, no question, but using it for anything more than inspiration ...

But I think if this thing is deployed widely it will crash and burn. It will rapidly get a reputation for leading people to disaster and that will be the end of it.

Re: GPT is all you need for the back end

#274

Post by ProllyInfamous »

bccdee wrote: post #267
ProllyInfamous wrote: post #258

Earlier quoted context omitted.

See also: Star Trek's Ferengi . If you are unfamiliar with Star Trek's "cosmoverse," just go to http://perplexity.AI , and ask it to explain anything from that storyline, including the Ferengi, who existed solely to allow the writers of Star Trek to "mock" some of human's more ridiculous notions of "fair" and "what is." It be happenin' The world isn't fair, and GPT-like technologies will and are already making sweepi…

I think the questions raised about creativity by generative neural nets have pretty simple answers. Creativity requires a certain amount of thought put into it: A writer doesn't just mash other pieces of writing together into their own work—they think about how they want the audience to react and what kinds of things they could do to create those reactions. An LLM doesn't have a model of an audience, nor can it even…

I have come to the realization that even among flesh, "creativity" is so rarely distributed as to be `inhuman`, IMHO.

Re: GPT is all you need for the back end

#275

Post by ProllyInfamous »

greenie_beans wrote: post #265
ProllyInfamous wrote: post #245

Earlier quoted context omitted.

Ask Perplexity.AI to explain what Kurt Vonnegut's "Timequake" (1997) computer program Palladio is capable of [spoilers follow]: >Here's the thing: Frank went to the drugstore for condoms or chewing gum or whatever, and the pharmacist told him that his sixteen-year-old daughter had become an architect and was thinking of droping out of high school because it was such a waste of time. She had designed a recreation cent…

> and killed himself the first time. is an oddly poetic way to say that. also, i tried getting chatgpt to list a bill of materials for a shed build and it refused. maybe one day.

RE: Shed.

You just need to ask the correct questions.

RE: Poetic Justness: Read TIMEQUAKE and it will being even sweeter, running through your brain the second time around...

Re: GPT is all you need for the back end

#276

Post by ProllyInfamous »

daveguy wrote: post #261
ProllyInfamous wrote: post #242

Earlier quoted context omitted.

>[using probability it] answers your question in the best way possible, based on the results of its training. In medical school, we were taught Differential Diagnosis, which is the manner which GPs MUST utilize to solve symptoms. This is a probabilistically-determined ranking of what is MOST LIKELY to be the cause, based on how the patient presents [medical symptoms]. A LLM like ChatGPT is already demonstrating can,…

> filter through many of the "first guess" diagnosis, and prevent unnecessary testing using information that extend beyond the single minds of patient and doctor. Source? I haven't heard anyone report this kind of capability for chatGPT or any other model. I have seen a lot of examples of very confident and wrong. Which is the opposite of what you want in a probabilistic assessment.

RE: Confident but wrong.

This is why the Human Interpreter (Dr. Prompt Engineer? Dreamweaver? Pattern Seeker?) is still so necessary.

For now, R.

Re: GPT is all you need for the back end

#277

Post by ProllyInfamous »

jostiniane wrote: post #270
ProllyInfamous wrote: post #240

Earlier quoted context omitted.

>...even the worst human programmers have error rates well below 1% FOR CODE THAT ACTUALLY SHIPS. I added that emphasis, and you should realize that shipping code is the result of teams, and engagement like ChatGPT is demonstrating will replace most of your fellow human teammembers, and the only human input may be "putting it all together..." this is a top-level jobtask, with limited employment opportunity; I worry t…

I have 3 main clients now: 1. project with large code bases that I maintain, the configs alone are of the scale of 20k lines of yaml files 2. Prototyping client with smaller code bases and varying requests for projects needing to be adapted for their clients 3. Implementing research results of a research institute In all what I have/saw, I still don't see a single use case for GPT assisted programming. Even for snipp…

You are an extremely sophisticated and well-trained crafts-person, obviously.

Now: IMAGINE ACCOMPLISHING THAT without ever having had been able to hold an "entry level position" at whichever creative-houses / apprenticeships helped in your current sophistry.

YES: If you are "top tier" educated and know how to communicate with these GPT systems... you can replace (e.g. as a lawyer) your entire law clerk staff, save one fast typist that is also a good question asker.

I have found in my previous six weeks of Prompt Crafting that it has helped me in so many areas of my life BEYOND TECHNICAL. For whatever-odd reason, of the smartest people I know, it's the most-technical that seem to have a problem with the rapid changing world of not needing 50% of your labor force for routine entry-level (and probably mid-level, too) for any job that currently just requires you to sit in front of a computer screen and not have any executive authority (literally, every single one of those jobs isn't "GONE," it's just that all the stuff that you currently pay younger people to do/learn-on-the-job... you don't need these people (correction: nearly as many of them) to chew through datasets. I am around 40, and have two lawyer brothers, and what they were able to accomplish in lawschool pales in comparison with what many ChatGPT systems I've seen and used and benefited from (again: outside of technical areas of expertise; OF COURSE IT GETS WRONG, so do humans; it is learning to ask better questions literally by the thousands... every second of every day.

The elitist mentality of "OH BUT MY JOB IS SECURE" is so silly, because what do you do when nobody is able to support themselves / help you. I often ask my clients: if Reagan's policies were `so good`, then how come it's `so hard to find good help`? This is a good question for which I have yet to ask AI, but it is a good one and probably not answerable in even just a few paragraphs.

Have a great day; be grateful for what you have; know that your social security (if in US) will be non-existant (for real, this time IS different) because it is already so out-of-whack that the massive unemployment issues WILL affect your life.

The previous few tens-of-thousands that a few FAANGs disposed of, pre-Christmas... is just the beginning folks!

Have a great workweek =P

Re: GPT is all you need for the back end

#278

Post by goatlover »

bufferoverflow wrote: post #271
goatlover wrote: post #92

Earlier quoted context omitted.

So when will these self-driving cars be ready for mass consumption? Because they're starting to sound like flying cars at this point. Yes, technically doable and there are always some working prototypes, but no real market, and not seen as practical.

We have flying cars.

Where can I purchase one to use on the road when I'm not flying (helicopters don't count as cars).
Post reply on HN