nave

joined 1 year ago
 
[–] [email protected] 9 points 3 days ago (3 children)

At the same time, o1 is not as capable as GPT-4o in a lot of areas. It doesn’t do as well on factual knowledge about the world. It also doesn’t have the ability to browse the web or process files and images. Still, the company believes it represents a brand-new class of capabilities. It was named o1 to indicate “resetting the counter back to 1.”

I think it’s more of a proof of concept then a fully functioning model at this point.

 

For OpenAI, o1 represents a step toward its broader goal of human-like artificial intelligence. More practically, it does a better job at writing code and solving multistep problems than previous models. But it’s also more expensive and slower to use than GPT-4o. OpenAI is calling this release of o1 a “preview” to emphasize how nascent it is.

The training behind o1 is fundamentally different from its predecessors, OpenAI’s research lead, Jerry Tworek, tells me, though the company is being vague about the exact details. He says o1 “has been trained using a completely new optimization algorithm and a new training dataset specifically tailored for it.”

OpenAI taught previous GPT models to mimic patterns from its training data. With o1, it trained the model to solve problems on its own using a technique known as reinforcement learning, which teaches the system through rewards and penalties. It then uses a “chain of thought” to process queries, similarly to how humans process problems by going through them step-by-step.

At the same time, o1 is not as capable as GPT-4o in a lot of areas. It doesn’t do as well on factual knowledge about the world. It also doesn’t have the ability to browse the web or process files and images. Still, the company believes it represents a brand-new class of capabilities. It was named o1 to indicate “resetting the counter back to 1.”

I think this is the most important part (emphasis mine):

As a result of this new training methodology, OpenAI says the model should be more accurate. “We have noticed that this model hallucinates less,” Tworek says. But the problem still persists. “We can’t say we solved hallucinations.”

[–] [email protected] 3 points 2 weeks ago

Weirdly it’s been working better for me.

 
 
 
[–] [email protected] 8 points 1 month ago (1 children)

For the first point:

“While he was in high school, Josh Shapiro was required to do a service project, which he and several classmates completed through a program that took them to a kibbutz in Israel where he worked on a farm and at a fishery,” Bonder told the outlet, and later confirmed with The Daily Beast. “The program also included volunteering on service projects on an Israeli army base. At no time was he engaged in any military activities,” Bonder added.

second: He did not say “savages” he said:

too battle-minded to be able to establish a peaceful homeland of their own

Additionally both of these statements were made in 1993 and he’s already stated he longer supports these statements. Source because you don’t have one.

Admittedly 3 is weird but without proof (which I didn’t find) it’s nothing more than a conspiracy theory.

 
[–] [email protected] 1 points 1 month ago (1 children)

I would like to enter for Crisis Core.

472
"no" (lemmy.ca)
 
[–] [email protected] 1 points 1 month ago

Well the songs that the YouTube videos mention are all instrumental so the lyrics aren’t necessary.

[–] [email protected] 3 points 1 month ago (2 children)

not good, let alone capable of passing as human-made music

I disagree. AI music is shockingly good at sounding like real music. Just a couple months ago a popular producer sampled an ai generated song without knowing it.

But soon after, it was revealed that the singer from the “BBL Drizzy” beat didn’t exist — the voice was AI-generated, as was the song itself. The vocals, melody and instrumental of the sample were generated by Udio, an AI music startup founded by former Google Deep Mind engineers. Though Metro was not aware of the source of the track when he used it, his tongue-in-cheek diss became the first notable use case of AI-generated sampling, proving the potential for AI to impact music production. (A representative for Metro Boomin did not respond to *Billboard’s *request for comment).

476
france (lemmy.ca)
submitted 1 month ago* (last edited 3 weeks ago) by [email protected] to c/[email protected]
 
[–] [email protected] 4 points 1 month ago

Hey I have that calculator too!

[–] [email protected] 1 points 1 month ago

Gaps #78 🎥 🟥 🟥 🟩 ⬛ ⬛ ⬛ https://gaps.wtf

[–] [email protected] 6 points 2 months ago* (last edited 2 months ago) (1 children)

Anbernic uses a Linux based os I believe.

[–] [email protected] 3 points 2 months ago* (last edited 2 months ago) (1 children)

It’s actually in Arizona.

[–] [email protected] 11 points 3 months ago (8 children)

Well then why did you describe them not doing that as malicious compliance?

view more: next ›