Large Bodied Male

CIA hater. OP hater. General hater.

[Large Bodied Male @CaryyJery]

#NewProfilePic

[Large Bodied Male @CaryyJery]

@exhales_deeply @JeffreyBiles Yea I could see it being more useful if money is ever less prevelant in AI. But right now you got a trillion dollars you're not spending it on Jev

[Large Bodied Male @CaryyJery]

@priestessofdada Ps. I'm getting my PhD for ml. This shit is all Chinese people. I've never met another white guy. They're all Chinese and they're the damn best. Every white guy I've talked to is a humbling bafoon

[Large Bodied Male @CaryyJery]

@priestessofdada To your first half, they were white. No one in AI research worth their salt is white. When a white guy is telling you how AI works he's full of it. AI researchers are Chinese. Not white. - Me, a white guy

[Large Bodied Male @CaryyJery]

@davidegallo @camhberg I didn't see it say they fine tuned the models anywhere

[Large Bodied Male @CaryyJery]

@danielshahryari @shauseth Bruh I just had GPT 5.6 screw up everything. I was so lazy I didn't read it and the 12 hour evolutionary search kept sucking dick. Eventually I read the code it straight up did NOTHING right. 1 hour to humanly fix it, 72 hours of my time wasted. This shits good.. sometimes

[Large Bodied Male @CaryyJery]

@yx3st @yuntiandeng How many times have I read that 😂 it's the moment I realize AI isn't taking my job

[Large Bodied Male @CaryyJery]

@AIDRIVR How the AI handles these edge cases is not simple imo. You either insert training data to ignore a stop sign or leave this data out and the model makes the same mistake again. Telling the model to ignore the stop sign could do more harm than good. Not easy

[Large Bodied Male @CaryyJery]

@lthlnkso Now calculate what portion of my stolen tax dollars are wasted on absurd fraudulent studies like this. Like seriously, who is being paid to produce a 200 page report of nonsense.

[Large Bodied Male @CaryyJery]

@MingM35169 @johnennis Learn to write some code like for realsies and do more than Claude. Until then, I like to know that a repo can be completely ignored.

[Large Bodied Male @CaryyJery]

@johnennis @itscoleonx Idk, sounds like it's more an author than you.

[Large Bodied Male @CaryyJery]

@AvneeshKasture @adkravetz @jconorgrogan Yea but that's not you argument. Your not talking about context window. Youre saying human language cap the intelligence of the models. If Einstein operated in this space, the models should be able to easily make conclusions as grand or grander than Einsteins, right?

[Large Bodied Male @CaryyJery]

@AvneeshKasture @adkravetz @jconorgrogan If only Einstein didn't have to operate in this space. Oh how far the world would have come.

[Large Bodied Male @CaryyJery]

@AvneeshKasture @adkravetz @jconorgrogan Agreed upon terms of art and science are used because that's how information is communicated effectively. The model could say the same thing without doing this nonsense. It's stupid. Annoying. It's part of the hype.

[Large Bodied Male @CaryyJery]

@AvneeshKasture @adkravetz @jconorgrogan Its indicative of RLHF tuning toward this cause people like you say OMG it's so heckin smart! It made up its own words to describe basic concepts!

For humans, I usually think it's indicative of an insecure and unknowledgeable con artist. It has nothing to do with intelligence.

[Large Bodied Male @CaryyJery]

@flowersslop Duh, that's literally how they work. Their input is literally supposed to be a noisy image. It basically did its job. If you added noise to the image, you actually just took the image back a few steps in the models output.

[Yun-Ta Tsai @yunta_tsai]

Many people think any given ML project is 99% training.

In reality, it’s 50% evaluation, 40% data cleaning, 8% integration, and 2% training.

The first two set the noise floor for learning. No ML magic matters; the model cannot lower the noise floor, as that’s the optimal bound of Shannon encoding of your data.

Thus, not a single day goes by without me thinking about ontology. Even the old labels have to be constantly reviewed.

[Elon Musk @elonmusk]

In the future, a trillion times a trillion dollars will be spent on making antimatter to travel to other star systems

[Large Bodied Male @CaryyJery]

@heynavtoor @scaling01 Is everyone else a bot, how did they not get this?? Though, a bot would probably get this too!

[Large Bodied Male @CaryyJery]

@trajan0x @gf_256 Lmao wut. Is there a reason for that?? Are they auto-generated?