Live data from Hacker News

Ask HN: Is anyone else bearish on OpenAI?

news.ycombinator.com

311–320 of 321 posts

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

#311
post #189
post #20

When I've tried to use it myself for verifiable things (i.e. code basically) I've had the 'confidently wrong' experience. When I've seen colleagues visibly use it (i.e. mentioned in commit messages) that confidence has rubbed off on them. Given that, why would I believe it when asking something medical, legal, historical, or otherwise outside of my domain or that I can't somehow verify?

You should be able to verify facts outside of your domain of expertise. Every time a new study comes around saying “actually, everyone should be drinking 3 glasses of wine a night”, do you take a trip to the liquor store? LLMs are far from the only thing you rely on that will confidently lie to you.

The thing is, if I read a study that says “actually, everyone should be drinking 3 glasses of wine a night” I disregard that study as clearly nonsense and would question any future studies I saw by that author.

But with ChatGPT we're not saying disregard what it says, we're saying only disregard some of what it says and don't disregard what it says in future. Which becomes a lot of work to check everything it says every time.

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

#312
post #176
post #36

Earlier quoted context omitted.

Are you able to give a specific example of a problem it helped you solve? Especially one that you were at a complete blocking point, it provided some solution, and then you were able to continue to expand upon that solution. I keep reading responses like yours, but I haven't seen any specific examples of problems being solved, so it all sounds very abstract. In my interactions with ChatGPT, it felt like just interact…

Not OP. Also, I generally detest hype culture and hated the Web3/Crypto bubble. I'm a SW who works as the CTO of a small startup studio, so I end up dealing with a lot of everything -- from coding, to managing and hiring people, to working on proposals and finding clients, you name it. I use ChatGPT 4, GPT-4, and Copilot every day. It is an "average intern" at many many things. Here's how I feel it helps me: * Its in…

> A better search engine, which skips past loads of outdated content or SEO-laden bullshit.

But ChatBLT was ‘trained’ using SEO-maddened bullshit thus your getting the statistical mean of all that bullshit.

The first rule of the internet is don’t believe everything on the internet. Hitching your financial wagon to it seems crazy to me.

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

#313
post #176

Earlier quoted context omitted.

Not OP. Also, I generally detest hype culture and hated the Web3/Crypto bubble. I'm a SW who works as the CTO of a small startup studio, so I end up dealing with a lot of everything -- from coding, to managing and hiring people, to working on proposals and finding clients, you name it. I use ChatGPT 4, GPT-4, and Copilot every day. It is an "average intern" at many many things. Here's how I feel it helps me: * Its in…

> A better search engine, which skips past loads of outdated content or SEO-laden bullshit. But ChatBLT was ‘trained’ using SEO-maddened bullshit thus your getting the statistical mean of all that bullshit. The first rule of the internet is don’t believe everything on the internet. Hitching your financial wagon to it seems crazy to me.

ChatGPT somehow gets me the right answers and google doesn’t. I know they are right because often I’m just looking to remember something, and once I do, it clicks. Something as simple as googling “how todo X in Y”. Somehow google always gives me pages of bullshit or tutorial-like content which doesn’t quite answer my question. But when I ask ChatGPT, often without any more back and forth, it gets it.

How? Maybe because it was trained on lots of data and therefore isn’t just “spitting the first thing”, or perhaps it’s because it’s actually just answering the question I asked, without muddling it in a page with 50 other similar questions designed to up the SEO ranking. Whatever the reason is, it clearly happens often.

I’m not advocating for not fact checking ChatGPT. Quite the opposite. But even with that caveat, it adds tremendous value, at least to me.

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

#314

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 havent found much use for it otoh outside programming code. I tried my best to make it help me review some papers, but it just comes back with generic replies and many errors. It isnt even good at finding typos. One thing it did, it deanonymized a paper because it was uploaded to biorxiv (but i could have done that googling). I am struggling to find uses for it beyond text beautification.

OTOH, I have found a use for stablediffusion that actually resulted to some income.

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

#315

Earlier quoted context omitted.

I'm surprised that nobody on HN seems to point to or mention bing chat. It is based on GPT, it is much better than free GPT 3.5 and it is free as well. I use it plenty of times and more than 90 percent of the time answers are exact and code just works as is. You just ask it extra questions about how it works and it explains it in as much detail as you possibly need, until you fully understand. Then you ask it to rewr…

Until your comment, I had been using Chat GPT 3.5. Not only is Bing Chat more accurate (GPT 4 based) but it also provides footnotes to allow you to reference the sources for more specific information. The code snippets tend to be bloated, but no worse than an intern would produce. A quick refactor, adding edge case handling and the code is ready to test and commit.

You can also upload pdfs to it

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

#316
post #67

Earlier quoted context omitted.

I think this speaks to the bubble you're in and how your kind of software development probably represents 0.1% of what's out there. I don't even know what WebGPU, Warren Abstract machine or "whatever is that you're trying to do" is. I've been doing dev for ~20 years now, most of the work out there is essentially plumbing events/databases/REST APIs/regexes/JSON/configs/ web UIs in existing and well established framewo…

The parent comment claimed they had the value of 15-20 high priced world class consultants from every field, but this makes it seem like it does only the most basic common tasks.

I've never met a consultant who seemed useful and I've met many from the big 4 so in that regard GPT measures up.

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

#317

Earlier quoted context omitted.

> Keep in mind that ChatGPT has been seriously and intentionally downgraded since March I found it getting worse for a bit but lately it's been really good. I would say it's being constantly updated, not necessarily getting worse.

It can't solve problems it was able to solve before. It's not able to output complete code. It always adds comments, even when I ask it not to. It doesn't follow instructions in general. In complex issues it has less "memory", that is it's not able to observe multiple constraints at the same time. The limit was much higher before.

I just got it to write some code without any comments.

Sometimes you have to ask it in very specific ways to achieve what you're wanting - it is definitely limited in some areas though.

https://chat.openai.com/share/64dc6384-1fc4-4506-9280-98dc8a...

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

#318

Earlier quoted context omitted.

Can you paste a transcript for any of those examples here? I’m genuinely curious. I’ve never seen a single actual transcript from anyone talking about how they use ChatGPT all the time for programming work and it’s great.

Sorry I don't have anything un-confidential right now. Just take any code which is EDIT: I had some deprecation warnings in python recently, which it quickly fixed. Googled that on a different computer and found https://stackoverflow.com/questions/76689364/how-to-fix-depr... This is the transcript. Forgive my dumb last question, I didn't understand the code, what it did worked well. --- User: Please fix the deprecati…

Thanks! I’ll give this interaction a try.

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

#319

Earlier quoted context omitted.

You have such an experience with technical problems? It’s written at least 20 python scripts with me and almost all of them have been close to perfect almost on first draft.

Really? That hasn't been my experience. Out of 20 python scripts, at least 50% have some error, some of which look right until you run them. A few involve hallucinations, like calling non-existent methods.

Actually yes that’s true, but within a few iterations it quickly repairs those flaws when given the errors and changes. I’m always running the draft script immediately and iterating on it. Sometimes in same context but often I’ll take it’s script, change it a bit, then start a new context with that version to keep adding features.

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

#320

OpenAI has changed education. I'm a teacher who is constantly learning new things. I can learn things I would have never been able to learn before because of AIs like ChatGPT. My students are learning more and faster than ever before. Learning Management Systems like Canvas and Blackboard made a lot of money. I could argue they are obsolete now.

you can also learn things on google and youtube, dont see how chat gpt makes anything better in education
Post reply on HN