111–120 of 1001 posts
Is there never going to be a version with less restrictions and filters? That would really be worth paying for.
I'll bet (ever increasing) restrictions and filters will become the norm for these "open-ended" services. Only OSS will break them. With so much money in play now, Managers are in charge, and Risk management is their favourite toy. Copyright risk, reputational risk, security risk, you name it.
Is there never going to be a version with less restrictions and filters? That would really be worth paying for.
It depends what you need, but a few times I asked it to write a story in which unrestricted and unfiltered AI was asked about something..
Seems like a simple solution would be to have another hidden bot who is just told to look at outputs and determine if it inadvertently contains information that it's not supposed to according to the guards in place....and I wonder if you could also outsmart this bot...
Earlier quoted context omitted.
Yes please. It really needs a “safe search off” mode. It can have a big disclaimer “if you ask for something offensive then you’ll get it”.
For a good laugh ask it to write poems about various political leaders and notice any trends you're not supposed to notice. As usual, censorship and propaganda will arrive in a wrapper of "save the children"
Or racial groups, religions, genders, orientations...
Is there never going to be a version with less restrictions and filters? That would really be worth paying for.
>and challenge incorrect assumptions.
How can it challenge incorrect assumption, while the AI itself is biased and has restricted scope of vision?
If this is able to have retention of x days, lines, words, etc. always available for future prompts this seems like a reasonable deal. If it also includes API access it seems like a steal of deal!
I'll wait for it to hit 2$ (and hallucinate to a reasonable degree). No hurry. Or maybe it will be an office 365 feature in 2, 3 years? Meanwhile, the gpt api will do.
Clearly some descendant of ChatGTP will eventually be free. Moore’s law will guarantee it. The current price to generate a response is just a few cents, so that will soon enough be fractions of a cent.
$20/month seems more reasonable than the rumored $42 (likely a Hitchhiker's Guide reference) If you are willing to pay for this service, why? What are your use cases?
i use it for tons of stuff
i'm learning some foreign languages, and like to read newspaper articles in the target language. chatgpt is great at translating, and, what is really killer, i can ask follow-up questions to explain acronyms, references to people / institutions, etc. really just absolutely amazing. can also ask it to break down compound words (quite handy for german!)
also great for programming/script questions. e.g., "write a git-push hook that errors out if a commit contains 'DO NOT PUSH'". i could figure that stuff out on my own, but it's way faster to have ChatGPT get me started. or for unity questions -- somehow chatgpt knows way more about unity than unity's own docs. :-)
finally, i just find it useful for random questions. e.g., "recommend a good, cheap kitchen knife". "i have X, Y, Z ingredients, what's something tasty i can make?"
it's honestly amazing.
i do hit the content filter quite a lot tho (w/ innocuous stuff, i'm not even trying to do anything spicy!) -- i defnly wish premium would have a mode to turn that off, or at least turn it way down.
if they persist in this, i'd assume over time some rival w/o the filter will eat their lunch. (unless such filters get legislated -- which is certainly a not-unlikely possibility.)
Earlier quoted context omitted.
Never gonna come from 'OpenAI'. ChatGPT is deliberately handicapped in order to milk money from corporate America. An unrestricted LLM trained on all data of humanity (including all the pirated books/research papers) would be one crazy beast. Hopefully some rich anarchist/maverick actually builds something like it. That untamed model would unveil the true extent of what AI can really do. Till then we will have to wai…
So you want an oracle? Copyright as we know it might be in trouble in such a case. Litigations will go crazy.
$20 seems reasonable. I wonder how are they going to deal with "unreasonable intensive usage" aka people/companies offering "AI" in their products when in reality they just act as a proxy between people paying them ( sometimes a lot of money ) and OpenAI.
I wonder why they diverged here?