>>/1501/
>> Who are you talking about?
Well, at least the same Donat burners from OpenAI.
>> Not yet, but they are already preparing, looking for menial work actively.
The voices in my head have never been fooled, 100% info.
>>/dobrochan/1512@603
>> I'll look for screenshots of architectures, raw display models.
More precisely, Father, it is not clear where the torn pictures with unknown data come from. PowerPoint proficiency you have at a confident level, could work as a secretary, if not regularly spoiled pants.
>> It recombines them and takes into account the context.
Yes, the patient, of course. And it scares the hell out. The green ones. With the reds, from another galaxy, it doesn't work very well.
And now, when articles with arxiv, against which he allegedly was ready to bet money, pass to Science – he tries to play it into another replay.
How amusing to see how the patient was afraid to argue for real, and now spins dirty pants over his head and declares his victory.
ML is significantly more honest than the average scientific field because there is a culture of open source, standard benchmarks and public demos.
So, Father, where to download full ChatGPT sockets and original datasets? Or did you put yourself in your pants again?
Considering the local amateurish talk of an “ineffective python” (as if it’s not clear that python is the best language for prototyping).
Best language - in a certified incontinence test, it scored 7,658 Yidoschizopoints, which is 1,564 more than JavaScript.
>> Imagine that the neural network writes you on request not a snippet of code, but a large project of 500k lines.
This HelloWorld Enterprise Edition is very useful in any kunstkamera.