I'm surprised anyone doesn't get value from them. I query GPT-4 regularly, and it rarely disappoints.

It's decent at coding and excellent at troubleshooting.

I've never seen it make a grammatical, syntactic or spelling error. With good context, it can write anything you want.

You need to have experimented with it a bit to know how to reliably prompt it, but once you get a feel for it the tool is invaluable.

I didn't mean to be rude or anything like that. You are probably right, seeing all these positive opinions, there must be value in there.

I'm just amazed by the reactions on this post. From the HN crowd I was, wongly, expecting very different opinions.

For coding, are you using it like a pairing with a junior developer or more like a search engine for documentation?

I use it with the documentation of what I'm building.

Paste in the relevant section of the developer docs, ask it to write a function, or ask it to produce a valid API request, etc.

I suppose it's like a junior dev in a sense but it's very fast, can correct its own mistakes if you point them out, and has an internet worth of knowledge about various approaches and troubleshooting techniques.

I find most people who don't see value haven't actually used GPT-4 yet. Have you?

I use LLM for generating data, for example to fill SQL tables for tests or some python datastructures. Transforming data works really well, so tasks like "take bibtex entry and convert it into citation"

I've tried to code entire IOS app with it but failed. You still need to have knowledge about what you are actually doing. I feel like it can accelerate learning, but in a very narrow way. When I was trying to do my own android app 3 years back I've learned bunch of different things about what I was trying to actually do and also around the whole ecosystem. When I was trying to code the IOS app I made progress fast, but I felt like I was just learning the thing that I wanted to learn not the actual surrounding knowledge. Now this sounds good, but I feel like I'm leaving something on the table when doing this kind of "accelerated learning".

I've also tried autogpt/langchain approaches to automate some writing commands for me like "remove all files that are beginning with abc". It did generate appropriate command but sometimes if they failed (the files weren't there in the first place) it just fell into the loop of constantly trying to get a successful deletion by refining the command. It even tried to go to the / to find all such files, talk about paperclips huh. Truth to be told I haven't played with the idea much so there is a room for improvement

LLMs kinda help, they are especially good when provided with source, but I don't really feel like a 1000x super-hacker with VR-glasses that spawns hundreds of agents each minute to do different task. (I would like to though)

I would like to do something with the LLMs but every idea I have already has an established startup or the argument "you could just write this in chatgpt" is made. Literally take whatever comes to your mind and add "LLM" or "GPT", there is a startup for that.

Lots of people use ChatGPT for things that hallucinations aren't an issue for. Drafting emails or articles or blog posts (which the human reviews and polishes) and other fun things like generating haikus are pretty common use cases.

Personally I also find the level of hallucinations (with GPT 4) to be minimal as long as it's a widely covered subject. The more technically niche, the worse the hallucinations get.

Many things. Writing tasks are aided but it helps a lot with coding imo. SO and the internet as a whole is more infuriating than helpful a lot of the time.

Also ChatGPT can do a lot of basic logic and repetitive tasks for me. It helps get things going and saves a ton of time and effort.

And that's not even mentioning GitHub Copilot.

Funny, I just cancelled my Copilot, I didn't really see the value in it.

It's fascinating how people see tools differently.

Autocompleting my console.log statements with uncanny accuracy is already worth $10 a month to me.

It's hit or miss so I'm not surprised to hear from someone who sees more misses than hits. But when it works well I feel this rush of gratitude that I don't need to type everything out.

I use it daily to write code for me, mostly of the "here is some input data, please use this API and this other API to transform it into this example output data". It works flawlessly 95% of the time.

I use it to search esoteric stuff about tools I don't use that much, and have horrible ux. What day is sunday to chron? How do I specify the first text field to 'cut'? How the heck do I do x thing in awk? gpt gives me an answer I can try and iterate on. I'm using it as a glorified search engine with an extremely clean and pleasant interface. I'm not asking it to write production code whole cloth.

TLDR: There's an old saying 'trust, but verify' and that applies to everything gpt gives you.

Have you actually sat down and used these models quite a bit?

I pay for GPT-4 and it is worth its weight in gold, I can engage in a conversation about technical topics.

Sure, they do "hallucinate" sometimes (so do my coworkers FWIW), but you can engage in a conversation about why it is wrong and they are quick to correct themselves (...this part can be less like certain coworkers).

Here's an example of a random question (relevant to my job) that I asked gpt-4

> In linear/logistic models, it is common to take the multiplicative interaction between two features which might be one hot encoded. For instance, A is a vector that one hot encodes your college and B is a vector that one hot encodes your employer, you would take the product len(A)*len(B) as the size of the new interaction vector. How do you make such cross/interaction terms in tensorflow?

GPT-4 is less full of shit than most people I know and I don't think it is even close.

I have always found the knock that it confidently answers things it really doesn't know even if wrong quite hysterical. That exactly describes some of the smartest people I have met in my life.

Depends on the level of experience of the person using it or the given task. Less experienced coders find it useful because it saves time spent learning how to do things. Less demanding tasks are trivial if they involve basic coding. For others it doesnt help much.

Bah! Programming is life long learning, GPT-4 is not just useful for beginners.

To be fair that’s true. If i switch from a language i have experience in to a language i am not familiar with then i find it useful. But i still dive into what its output does so i can actually learn the new language.

The company I work for uses ChatGPT, StableDiffusion, ElevenLabs, and lots of other AI tools every day now. We make ads and use it to generate scripts, create "stock" photos, fake voice over, etc. We even have started training our own models to match branding guidelines and whatnot.

I also use ChatGPT and SD to help me with my personal hobby of game development. It's good at analyzing code snippets, refactoring, and SD is good for inspiration.

These are really useful tools and it's only the beginning.

You can paste some gnarly code into gpt4, tell it "I think this code has an off by 1 error in it someplace" and GPT4 will find it for you.

So yes, it is useful.

It has helped me many times, "You've listed the items in this config file in the wrong order", saved me hours of debugging right there.

Even stupid stuff that I do once every couple years, I can ask GPT to setup my initial express server stuff "I want CORS enabled and these endpoints to process JSON" and it pops out code a few seconds later and saves me maybe 5 or 10 minutes of Googling.

I recently had GPT4 write me skeleton code for using Websockets, I'd never used them before and having something to base my work off of saved me a lot of time.