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

Ask HN: Is anyone else bearish on OpenAI?

news.ycombinator.com

Post reply on HN
Print view View original

41–50 of 321 posts

Re: Ask HN: Is anyone else bearish on OpenAI?

#41

Post by alisonatwork »

bravetraveler wrote: post #6

I am, personally... at least until it's all done on-device and functions offline. I have no trust with the siphoning of data we see at large I suspect I'm fairly alone on this. They'll probably do well without me. Most people that even know about it probably don't mind. I can't even verbalize why I do

This is basically it for me.

I've had a play around with some OpenAI-powered sites and it is neat how much it is capable of, but I feel uncomfortable typing personalized prompts or detailed questions into a system where I know everything I type is going to be harvested. You could argue that by commenting on HN or posting anywhere in the internet everything I type is also going to be harvested (perhaps into the very same models), but that contract was always clear. There is a difference between companies using information I have chosen to share publicly, and companies doing the same with what is presented as a private exchange.

But once they can fit that mini GPT into my pocket, and the learning it's doing is truly personalized to my own install... for me that will be a much more appealing product. I guess the technology will get there, eventually.

Re: Ask HN: Is anyone else bearish on OpenAI?

#42

Post by 2OEH8eoCRo0 »

eachro wrote: post #18
ghshephard wrote: post #9

OpenAI, at least in my day-day workflow for the last 9+ months has so superseded anything that google ever was to me that I'm having a difficult time comparing the two. I've got a monitor dedicated 100% of the time to ChatGPT, and I interact with it non stop during the flow of technical scenarios and troubleshooting situations that flow into me - working in areas that I have the slimmest of backgrounds in, and shutti…

How do you control for hallucinations?

Ask for sources. Ask if what it has already told me makes sense. Knowing that hallucinations happen and not to blindly trust.

Learn the limits of your tools.

Re: Ask HN: Is anyone else bearish on OpenAI?

#43

Post by theossuary »

aprdm wrote: post #32
theossuary wrote: post #19

Earlier quoted context omitted.

This interesting, I've had the opposite experience so far. I tried to do that because if it worked it'd be game changing; but it seems every question I ask it, I get half truths and lies back. Its impossible to figure out what is and isn't true either, because the LLM says everything with such a certainty. I end up having to research everything it says again to figure out what to trust. Sometimes it's good to get a f…

Are you using 3.5 or 4? And what kind of things are you trying to do ? It's not like asking it, copy + pasting results + paste errors loop takes a long time to validate if it works or not. Even when it doesn't work it saves me time vs google-ing and also not solving my problem. Or gives me insights.

In all honesty I've only been using InstantClaude so far due to the larger context window being something I'd want down the line. What should I be using in your opinion to get the best results?

Examples I can think of where it gave wrong info was:

- I asked to compare the size of the MacBook m1 and the 16" framework laptops

- I asked how to add the nixpgs-unstable channel in NixOS.

- I asked to help write a power shell script to create AD accounts.

Re: Ask HN: Is anyone else bearish on OpenAI?

#44

Post by aprdm »

floxy wrote: post #39
alchemist1e9 wrote: post #26

> What am I missing here? I’m guessing you haven’t actually been using it personally beyond some superficial examples. Once you use it regularly to solve real world technical problems it’s pretty huge deal and the only people so far that I’ve met who voice ideas similar to yours, just simply haven’t used it beyond asking it questions which it isn’t designed for.

I wonder if anyone has a favorite pointer or two to their favorite real world examples.

> I am using Linux and trying to set up a systemd service that needs X to display a UI through VNC, how can I get the X authorization token in my systemd file

> I'm using python and my string may contain file paths inside, for example: (...) , For anything that looks like a filepath inside the string we should replace its full path, for example (..)

> Can you write me a python script to kill processes in Windows that no longer belong to a process named "gitlab"

> I want to write a test for it that mocks both sentry_sdk and LoggingIntegration so that I can give my own mocks during test

> I want to create a Python script

It should be able to be called from the command like, like below (example)

Write me the script

;;;

All real examples from last week that took me 1 minute to be solved instead of of googling or creating from scratch / thinking about it

Re: Ask HN: Is anyone else bearish on OpenAI?

#45

Post by soasdfg »

ghshephard wrote: post #9

OpenAI, at least in my day-day workflow for the last 9+ months has so superseded anything that google ever was to me that I'm having a difficult time comparing the two. I've got a monitor dedicated 100% of the time to ChatGPT, and I interact with it non stop during the flow of technical scenarios and troubleshooting situations that flow into me - working in areas that I have the slimmest of backgrounds in, and shutti…

I've also gotten good results that have added value to whatever I was working on at the time, but I also get a lot of partially correct responses, or responses where ChatGPT will forget what I told it seconds ago and repeat the same incorrect response from a few prompts ago. It's very effective if you know enough about what you ask it to disregard incorrect information quickly.

It can be very helpful to guide you in a direction maybe you didn't consider looking into to begin with, but I don't fully trust it.

There was a fun example I had a few months ago where I wanted to see how well it would do with being asked to solve a problem iteratively instead of recursively, something like: "generate all valid, distinct permutations of a provided string given that you have a dictionary to check valid words against" and it got most of the problem correct, but when asked to fix anything it would go right back to a recursive solution with the same issue appearing, or in some cases a new issue.

It got me most of the way there with some edge cases I needed to handle myself, but it definitely seemed like that was as far as it was going to be able to go

Re: Ask HN: Is anyone else bearish on OpenAI?

#47

Post by mywacaday »

ghshephard wrote: post #9

OpenAI, at least in my day-day workflow for the last 9+ months has so superseded anything that google ever was to me that I'm having a difficult time comparing the two. I've got a monitor dedicated 100% of the time to ChatGPT, and I interact with it non stop during the flow of technical scenarios and troubleshooting situations that flow into me - working in areas that I have the slimmest of backgrounds in, and shutti…

I'm bullish in the short to medium term. I also pay the $20 a month as it's worth it for the quick in co text answers. The company I work for (60k+ employees) has also rolled out an internal version Of ChatGPT and will be integrating live data with our ticketing system which will be a game changer from the demos I've seen. We are also doing a 9M trial of office copilot which I think that will be as big as office 365 as it's $30 per user per month which is costing 32 per user per month. Also bullish on Microsoft as a result. Hard to see how MS will be out manuvered in the enterprise.

Re: Ask HN: Is anyone else bearish on OpenAI?

#48

Post by dangus »

throw111123 wrote: post #33

I think both Bitcoin and ChatGPT are revolutionary. I use Bitcoin regularly, because I live in a third world country where it's really hard not to get your salary seized. I use ChatGPT every day for lots of things and it has replaced Google search for me. And StackOverflow, of course. Notice how I said BITCOIN and CHATGPT. Not "crypto" and "ai".

Well, regarding your last sentence, I think that’s what OP is getting at.

ChatGPT made one great product. Bitcoin is one great “product.”

But the successors in the same category aren’t doing anything wildly more useful than the original “killer app.”

Crypto never really revolutionized finance, it just provided one solid digital currency product. Smart contracts and NFTs went pretty much nowhere and I struggle to identify a way that they are used in a widespread manner.

You and I are using ChatGPT regularly and it helps us quite a lot, but it hasn’t revolutionized life nor has it turned me into a 10x developer or something like that. It’s a service that is collecting $20 a month and that’s about the extent of its economic value so far.

In other words, “Replacing Google and Stack Overflow” is arguably not that exciting.

(Then I end up going back to SO/Google when ChatGPT tells me shit that is wrong)

I do think LLMs have way more potential than “crypto” but it remains to be seen how much more that is

Re: Ask HN: Is anyone else bearish on OpenAI?

#50

Post by enumjorge »

ghshephard wrote: post #9

OpenAI, at least in my day-day workflow for the last 9+ months has so superseded anything that google ever was to me that I'm having a difficult time comparing the two. I've got a monitor dedicated 100% of the time to ChatGPT, and I interact with it non stop during the flow of technical scenarios and troubleshooting situations that flow into me - working in areas that I have the slimmest of backgrounds in, and shutti…

I'd pay money to watch someone like you work. Like other responders, I've tried using chatgpt multiple times, and very often when I've used it for tasks where I was familiar with the subject matter, I was disappointed in the results. It's possible I haven't learned the proper way to get what I need from it and would love to learn how others are using it to solve real problems.

I mean it. I'd pay money to watch streams of someone using chatgpt to solve non-trivial problems.

Post reply on HN