jameis 3 picks
god im so happy watson has his achilles shredded
So anyway, no. I donât think generative AI is actually good at summarizing anymore.
he really has never changed, huh? dude will never not be an interception throwing machine. a machine that turns passes into intercepted passes
REAL ball dawg tears of joy are being shed
take that rapist behind the shed and do em like yeller but its a yeller that was an asshole and bit people for no goddamn reason
The text reflects a critical perspective on modern language models like ChatGPT. The speaker, who has studied machine learning extensively, notes that the core algorithms behind these models are decades old and remain largely unchanged. While they understand how these systems work, they argue that the stochastic nature of output generation (selecting likely next words) gives the illusion of synthesis and coherence without actual understanding.
They express concerns about the harmful effects of separating users from original sources, the reinforcement of biases, and the authoritative tone of the models. They also highlight the risk of propagating outdated or incorrect information if the training data contains long-standing inaccuracies. Overall, the text critiques the reliability and limitations of language models in producing accurate, synthesized information.
well. technically the rapist isnât uniquely identifying here
will the 30-30 season ever happen again
i dont think a qb will ever have as long a leash to get away with a 30/30 season ever again
anthony richardson might be the next member of the club tho
fwiw i find this to be misleading. i was trying to explain that iâm not just a blind skeptic or being a Luddite, but rather that i have engaged with on a scholarly level. i was not trying to come across like an expert which i feel like this implies more than it should
1 Like
yeah well one isnt on a 230 million dollar contract and is fun so
i think the ai misgendered nbow if i remember right
1 Like
whats the matter? we canât both be worthless hacks.
Itâs not ur fault I just like blaming may
Sorryyyyyy i donât mean cause you more argue. Please dont hate me u arent disappointed at me right? Please dont hate me please dont hate me nyuu huuuuuuu
Also itâs not that I think it will always make mistakes, itâs that it will way more than it should to really be more useful than doing it yourself. If itâs accurate 95% of the time, and you make 20 flash cards, one of those will be inaccurate on average. If you had spent the 10 minutes making the flash cards, youâre gonna be accurate basically always, and making them also helps you with retention.
1 Like
Youâre fine.
1 Like
Yeah idk I feel like youâre harming yourself long term with this. Being good at sifting through search results for the answer youâre looking for is important for CS/IT. There was an error I had to debug once that required me to find a particular page of a book on Linux that described an esoteric use case for a signal, and then had to dig into a toolâs source code to understand that yeah, the thing I was using could cause that signal to be thrown in theory.
But fuck it, letâs put it to the test. Ask ChatGPT what it means when a C program throws a âbus errorâ.
2 Likes
i feel like a lot of people just use chatgpt as google when you could just Use Google. why ask chatgpt a question when u could type the exact same thing into a search bar and like. See where the info is from.
1 Like
Yeah, and regarding my example, when the answers were posted actually helped point me in the right direction. Thereâs a wrinkle that Iâm curious if it will even include, but Iâll be flabbergasted if it puts that detail in proper temporal context.
1 Like
Thing that drives me crazy is I canât guage confidence on it for most use cases. Itâs functionally impossible to get an âanswer DNEâ. Google I can go to lots of different sites and see if the answers agree
2 Likes