Sign in

Lovah Spaghets!

@spaghets.bsky.social
71 followers 38 following 76 posts
PostsRepliesMedia
Lovah Spaghets! @spaghets.bsky.social · 23/09/2026
"In these “latent spaces” concepts become directions, conceptual categories become clusters, and reasoning unfolds through mutually informed transformations of sequences of high-dimensional vector patterns." Those directions in that space are knowledge about the world.
000
Lovah Spaghets! @spaghets.bsky.social · 23/09/2026
LLMs process representations in high-dimensional vector spaces where meaning is encoded in geometry. aiprospects.substack.com/p/llms-and-b...
aiprospects.substack.com
LLMs and Beyond: All Roads Lead to Latent Space
Essential concepts for understanding AI today and prospects for tomorrow. (Bonus footnotes: the maths of exponential orthogonality & updated citations 30 May 2025)
100
Lovah Spaghets! @spaghets.bsky.social · 23/09/2026
It's literally how LLMs work.
200
Lovah Spaghets! @spaghets.bsky.social · 22/09/2026
Busing doesn't help this. These kids are way behind even when they're in the exact same classrooms as kids who are doing well.
010
Lovah Spaghets! @spaghets.bsky.social · 22/09/2026
It's your argument. 🤷 Is producing unreliable output an argument for or against AI behaving like humans?
100
Lovah Spaghets! @spaghets.bsky.social · 22/09/2026
And I think if we could roll back time a few minutes, your brain would do the exact same things it did the first time. I'm not even sure what argument you're making.
000
Lovah Spaghets! @spaghets.bsky.social · 22/09/2026
>They work exactly according to their algorithm. So what? Why do you believe an algorithm is incapable of thought? Yes, autocomplete can think if it's backed by a sufficient process. You're just restating the Chinese room argument.
000
Lovah Spaghets! @spaghets.bsky.social · 22/09/2026
You say that like it matters at all. How does knowing what each node is doing exclude their collective state being able to infer facts about the world?
100
Lovah Spaghets! @spaghets.bsky.social · 21/09/2026
Evidence of what, exactly? These models are unquestionably intelligent, are capable of reasoning about problems that humans can't solve, create art and music. This is reality today, and anyone who uses them knows it. And I got my CS degree in 1996.
110
Reposted by Lovah Spaghets!
Will Stancil @whstancil.bsky.social · 21/09/2026
One thing I'll say: a whoooole lot of people are staking out strong philosophical views on AI with a false sense of security, serenely unaware that the capabilities that would severely undermine their confident claims already exist.
2422510
Lovah Spaghets! @spaghets.bsky.social · 21/09/2026
The fact that you're admitting that humans have the exact same failing as the ai - hallucinations - as an argument against the AI is very interesting.
100
Lovah Spaghets! @spaghets.bsky.social · 21/09/2026
Except it the exact same argument, just used in a way you don't like.
150
Lovah Spaghets! @spaghets.bsky.social · 21/09/2026
But not via reasoning and induction/deduction.
110
Lovah Spaghets! @spaghets.bsky.social · 21/09/2026
So what? It still solved the problem when no human was able to and not just by trying every possible combination or something equally boring.
110
Lovah Spaghets! @spaghets.bsky.social · 21/09/2026
When the prompt is "solve this unsolved math problem" and it does it, that's amazing and not just repeating the question back.
070
Lovah Spaghets! @spaghets.bsky.social · 21/09/2026
This is not brute forcing like they do when trying to crack large primes. This is more like hiring 1000 mathematicians to work on the problem at the same time.
110
Lovah Spaghets! @spaghets.bsky.social · 25/02/2026
You're using an opinion piece to argue the science?
000
Lovah Spaghets! @spaghets.bsky.social · 10/12/2025
Woah! Amazing!
000
Lovah Spaghets! @spaghets.bsky.social · 27/05/2025
Oh well! www.sanluisobispo.com/living/artic...
sanluisobispo.com
Warning Signs for Tesla: New Model Y Struggles to Find Buyers
Tesla’s much-anticipated Model Y refresh isn’t generating the buzz many expected, leaving the EV giant scrambling for answers.
010
Lovah Spaghets! @spaghets.bsky.social · 27/05/2025
Ai art, and AI in general, doesn't need any defense. Almost everyone is using it every day for work, and that's not going to stop. Chatgpt is the 5th most used site on the Internet and rising.
030
Lovah Spaghets! @spaghets.bsky.social · 27/05/2025
The AI industry won't be killed by this, it will simply become the Chinese AI that everyone uses.
000
Lovah Spaghets! @spaghets.bsky.social · 27/05/2025
This is how you lose your user base to a competitor who uses AI.
000
Lovah Spaghets! @spaghets.bsky.social · 17/03/2025
Sushi.
000
Lovah Spaghets! @spaghets.bsky.social · 17/03/2025
Buying a Tesla has never been more completely out of the question.
000
Lovah Spaghets! @spaghets.bsky.social · 07/03/2025
If it's a fake issue, then why are the democrats in the Senate so out of touch with the members of the Democratic party on the issue?
000
Lovah Spaghets! @spaghets.bsky.social · 07/03/2025
Certainly not. www.researchgate.net/figure/Compa...
researchgate.net
Figure 1 — Comparison of the number of medals won by Kenya and Ethiopia...
Download scientific diagram | — Comparison of the number of medals won by Kenya and Ethiopia versus rest of world in men’s middle- and long- distance events (800 m to marathon) in Olympic Games from M...
000
Lovah Spaghets! @spaghets.bsky.social · 21/02/2025
Screaming "LITERALLY" doesn't make it so.
100
Lovah Spaghets! @spaghets.bsky.social · 06/02/2025
Indigenous knowledge is a perfect example of a stopped clock being right twice a day. It's mostly useless.
000
Lovah Spaghets! @spaghets.bsky.social · 22/01/2025
But does Erin identify as one?
140
Lovah Spaghets! @spaghets.bsky.social · 19/01/2025
Which part of this was violated? There are still dozens of social media apps, including the one you're currently on, where you can speak freely. Part of the supreme court's decision was that this was a content neutral ban - it had nothing to do with what people were posting.
000
Reposted by Lovah Spaghets!
Jerry Coyne @evolutionistrue.bsky.social · 19/01/2025
A consortium of secular organizations refuses to recognize that in a few cases (e.g., sports, hormones for kids), trans rights conflict with ethicality or the rights of other groups. The Center for Inquiry begs to differ. I suggest supporting the CFI. whyevolutionistrue.com/2025/01/19/c...
1124
Lovah Spaghets! @spaghets.bsky.social · 04/01/2025
Lol. Not going to happen.
000
Lovah Spaghets! @spaghets.bsky.social · 31/12/2024
That is the case nearly all the time. Nobody cares that someone wants to play out whatever gender stereotype they like. The only time it's a problem is when they feel the need to invade female only spaces.
100
Lovah Spaghets! @spaghets.bsky.social · 30/12/2024
Follow the science...but not there!!
010
Reposted by Lovah Spaghets!
Terminally Online Leftist 🥂 @terminallyol.bsky.social · 23/12/2024
1344447716338
Reposted by Lovah Spaghets!
Michael Shermer @michaelshermer.bsky.social · 22/12/2024
Wokeness Poisons Science Part II of my series on why I am no longer woke open.substack.com/pub/michaels...
open.substack.com
Wokeness Poisons Science
Part II of my series on why I am no longer woke
5156
Lovah Spaghets! @spaghets.bsky.social · 20/12/2024
Woah! arcprize.org/blog/oai-o3-...
arcprize.org
OpenAI o3 Breakthrough High Score on ARC-AGI-Pub
OpenAI o3 scores 75.7% on ARC-AGI public leaderboard.
010
Lovah Spaghets! @spaghets.bsky.social · 20/12/2024
Exactly! It's a tool, the same as any other. It can be used for copyright infringement just as easily as a photocopy machine, but that doesn't make the machine the problem.
010
Lovah Spaghets! @spaghets.bsky.social · 20/12/2024
So why did you commit copyright infringement by querying for it?
110
Lovah Spaghets! @spaghets.bsky.social · 20/12/2024
No, there aren't at all. You don't have an even basic understanding of how LLMs work.
000
Lovah Spaghets! @spaghets.bsky.social · 20/12/2024
Literally any stream of ones and zeros can be easily transformed into any copyrighted work. The ones and zeroes aren't the important part. The end product is.
000
Lovah Spaghets! @spaghets.bsky.social · 19/12/2024
That is exactly how the law works. You can't copyright raw data, and that's ultimately all the a llm is. It's the expression of that data that can run afowl of copyright. The llm is not the thing violating copyright, it's the person writing the prompt.
100
Lovah Spaghets! @spaghets.bsky.social · 19/12/2024
If there's no copy of the work being stored, there's no copyright violation. That's what the case will come down to, and the AI companies will win.
000
Lovah Spaghets! @spaghets.bsky.social · 19/12/2024
It's different from humans, but I don't see how that matters at all. They don't actually copy anyone's work - they simply make correlations between words and ultimately concepts and higher order relationships and use those relationships to produce new art. It's a new thing altogether.
100
Reposted by Lovah Spaghets!
Adam Kinzinger @adamkinzinger.substack.com · 18/12/2024
If you reject 100 percent of conspiracies, you will be right 99 percent of the time . If you believe them all, you will be right 1 percent of the time.
25682951241
Reposted by Lovah Spaghets!
Lay ✨ @laythick.bsky.social · 17/12/2024
I’m making healthy eating choices in 2025. Goodbye 10pc wings and hello 8pcs wings
1994425296
Lovah Spaghets! @spaghets.bsky.social · 18/12/2024
www.youtube.com/watch?v=7OnR...
youtube.com
Donald Trump pooped his pants in France ahead of Notre Dame reopening!?
YouTube video by Mind-Bend Moments
010
Lovah Spaghets! @spaghets.bsky.social · 18/12/2024
I'm confused. Is 10 out of 510,000 important enough to focus on or not?
150
Lovah Spaghets! @spaghets.bsky.social · 18/12/2024
Yep. I certainly don't need any very special episodes in my honor.
120
Lovah Spaghets! @spaghets.bsky.social · 17/12/2024
Why would they make a whole episode about someone so completely irrelevant, then?
340