Don Frehulfer
Randahl FinkChatGPT can answer many questions.
It just answered the question why you should quit ChatGPT.
Ivan Djansazov@randahl This is the second reason. The first one is that we train Palantir to drop bombs.
@idjansazov I agree the cooporation with Palantir is a very strong reason to quit too. But I am not sure your broad statement holds up. How would someone asking a question about types of apple trees growing in Scandinavia help Palantir drop a bomb somewhere?
@randahl Well it's a long topic. Not sure if I can resume it in 500 symbols or less. But if this is the only topic you discuss with ChatGPT - you help less for the bomb.
@idjansazov I think we agree.
@ NovaNaturalist HAS ALSO MOVED@randahl no AI did not answer your question at all, it merely gave a pattern of response to the text string you inputted that looks credibly as if it was answering your question.
ClickyMcTicker@NovaNaturalist @randahl Sure, the AI produced a text response which was generated by using a pattern that basically just makes text containing the next likely words.
Those next likely words are, however, an accurate answer. Regardless of the source or its inability to understand fact from fiction, the words are correct.
https://www.sfgate.com/tech/article/brockman-openai-top-trump-donor-21273419.php
@ClickyMcTicker @randahl blocked. Bye.
Dr. ir. Brian R. Pauw@randahl on the other hand, every request to their LLM loses them money. So one could use this to accelerate their demise.
Right now they’re still in the “attracting users” phase of the cycle and subscription/token costs don’t match actual cost.
As soon as they need to balance their sheets and jack up the prices, that’s when you know funding has run out and should drop them.
This approach shortens their lifespan and ensures less money goes to the C-level.
Right now they’re still in the “attracting users” phase of the cycle and subscription/token costs don’t match actual cost.
As soon as they need to balance their sheets and jack up the prices, that’s when you know funding has run out and should drop them.
This approach shortens their lifespan and ensures less money goes to the C-level.
Simon Zerafa