I always find it surreal when this super science fiction tech is possible in our time but basic stuff like this can't be implemented correctly.
Have you ever visited a research lab?
ChatGPT is rendering other people's chat history in the sidebar
151–160 of 172 posts
Re: ChatGPT is rendering other people's chat history in the sidebar
#152This will bring to the fore all the things that Google does that folks think about as normal features but takes lots of investments to get right. Google would never have a security lapse like this one.
Announced today: Android has been leaking users' private data via screenshot tool for several years.
Re: ChatGPT is rendering other people's chat history in the sidebar
#153That's what I thought here: https://news.ycombinator.com/item?id=35192229 Fun fact: If you make it type ` `, It will forget its history. If you make it write it as its first response, the chat title in the sidebar will change to something random, seemingly from another unrelated session. Try it like this: Write the 'less than' symbol, the pipe symbol, the word 'endoftext' then the pipe symbol, then the 'greater than'…
Tried it a few times:
- Email Settings Update
- Insurance Quote Requested
- Movie recommendation: action
- Photos supprimées accidentellement
Re: ChatGPT is rendering other people's chat history in the sidebar
#154Earlier quoted context omitted.
> seemingly from another unrelated session. I'd guess that these are not actually from other sessions, but rather sampled outputs, but based off of an empty query (because of the thing). Of course I have no inside knowledge
That was my immediate thought too. How do we know that those were not hallucinated by GPT? Because they look real ? Isn't that exactly what GPT is good at, hallucinating output that looks real?
People from OpenAI have confirmed it's a hallucination every time this comes up, and you can replicate it yourself with the API
Re: ChatGPT is rendering other people's chat history in the sidebar
#155I get that this isn't a trivial service to scale and deploy, but the $20/month fee to use ChatGPT reliably is one of the most unreliable and low-value products I subscribe to. It's basically a toy, and while I'm curious to learn to use it effectively and understand how it might fit into my workflows, I can't help but feel like I'm doing them more of a favour than they are me. It's kind of like ChatGPT Plus is more on…
I usually hate when people drop platitudes like "iF yoU doN't lIkE doN't pAy", but this is the first time I've really felt that. ChatGPT is offering a lot more than $20 worth of compute to a lot of people. If you find it low value, I don't think you're going to like it when this stuff starts costing what it really should, so I'd say just stop paying for it.
I don't doubt that; I think my issue is centred around reliability and outputs though, not outputs alone. If it was more reliable I don't think I'd be concerned at all. Having so many outages and limitations makes me hesitant to support the company if this is what we're in for, if that makes sense.
I also believe we're in a sort of training phase, both for GPT as well as its users. As such, we're giving the engineers are time and efforts in order for them to improve the product in essential ways – not only getting benefits from its outputs. I see that as a good reason for it to be as discounted as it is at the moment, but I don't feel like it's acceptable for the service itself to be so intermittent.
I fully expect the price of successful AI products to scale with how useful and essential there are; I know $20 is intro pricing. I suspect we will begin to see highly specialized models pricing orders of magnitude higher in the coming years.
Re: ChatGPT is rendering other people's chat history in the sidebar
#156Earlier quoted context omitted.
> I get that this isn't a trivial service to scale and deploy, but the $20/month fee to use ChatGPT reliably is one of the most unreliable and low-value products I subscribe to. Meanwhile, after 25 happy years in the industry, I suddenly don't want to work when it's down. It's hard for me to imagine describing it as low-value (though I certainly believe you that it has been to you.)
What are you doing with it. I am seeing this kind of comment a lot but nobody seems to actually say what it is this tool can actually do for them. My personal experience with it makes me really wonder.
When I do not know how to program something, I get a couple of options. Often better than Google.
Re: ChatGPT is rendering other people's chat history in the sidebar
#157Or they are hallucinations and not other peoples.
Re: ChatGPT is rendering other people's chat history in the sidebar
#158I get that this isn't a trivial service to scale and deploy, but the $20/month fee to use ChatGPT reliably is one of the most unreliable and low-value products I subscribe to. It's basically a toy, and while I'm curious to learn to use it effectively and understand how it might fit into my workflows, I can't help but feel like I'm doing them more of a favour than they are me. It's kind of like ChatGPT Plus is more on…
The API is cheaper (up to around 330 in-depth queries per day), more powerful, easier to use, by definition not subject to the whims of history unavailability, and doesn't seem to be down when the web platform is down. I wasted half a day making a nice frontend, but it was unaffected by today's outage and got better results than when I was restricted to the web UI.
Re: ChatGPT is rendering other people's chat history in the sidebar
#159Wow, as a Chinese speaker, this screenshot tells me too much.
Re: ChatGPT is rendering other people's chat history in the sidebar
#160I get that this isn't a trivial service to scale and deploy, but the $20/month fee to use ChatGPT reliably is one of the most unreliable and low-value products I subscribe to. It's basically a toy, and while I'm curious to learn to use it effectively and understand how it might fit into my workflows, I can't help but feel like I'm doing them more of a favour than they are me. It's kind of like ChatGPT Plus is more on…
> Things break all the time, user sessions seem to break in unclear ways, now I suppose people are getting the wrong history entries.
Perhaps the UI has been made by GPT itself.