twobitshifter

Why did Google give Android away for free?

Look at pi coding agent.

As an open weights model Kimi will quickly find hosts other than moonshot, so I don’t expect this to really cause any issues for open weights adoption.

distillation attack? why the violent word choice? When OpenAI crawled Github was that an attack?

My parents’ ‘80s Chrysler New Yorker would talk instead of beep. ‘A door is ajar,” it would say, and we would giggle, ‘no, a door is a door!’ After Kitt in Nightrider, car companies wanted to be cool.

https://youtube.com/shorts/IhSrmB2qCdY?is=vg47_htvXpXS2ElU

That’s a figure mistake, it’s a figure from later on where they replace mars with earth in the J space, and when they do that, they see the model answers blue. This indicates that the reasoning is happening within the J space.

If we keep pardons, no reason to give that power to the executive branch at all.

Taking a view from outside the USA, European companies just had Fable taken away due to US export controls, and before that Anthropic announced it is holding their data for 30 days. There is immediate value to these firms to build their infrastructure around an AI that won’t be pulled away from them. And outside of Europe, other countries are more price sensitive and don’t have the same fear of building relationships with Chinese companies.

Probably the data center where the model is running more than anything. Another option is if Opus is using anything like a Mixture of Experts approach, in which case the amount of the model loaded in memory at one time could be smaller than GLM.

Right, just give the text llm access to a vision specific agent and that problem can be solved. Or if you really want let it even call Opus with an image - seems like you’d still save money

It works in a plane, so it’s somehow able to detect the higher rate of speed, maybe by integrating, but I think it does use signals from CarPlay as well.

Websites block certain throwaway email domains from signups. The concern is that this will happen with private.icloud.com

A good example of a throwaway email that is now useless because of these blocks is mailinator.com. Originally, you could just make up a random email on the spot like gregsrightfoot@mailinator.com, visit mailinator.com, and get the needed signup verification email. These services autodeleted messages and required no signup so they were a black hole for spam. However websites eventually got wise that their spam wasn’t being seen and started blocking the domain. Mailinator came up with alternative domains and there was a brief back and forth before the throwaway email domains all ended up being blocked.

Might turn out like when the nba banned air jordans.

Most popular music we hear outside is based on guitars and pianos, as I play the guitar I always think of music that way. You are right that we naturally sing in just intonation and woodwinds have difficulty with equal temperament based on their physics, however it’s very rare that you will hear a specific song in just intonation. You can search YouTube for performances, it’s pretty interesting to hear the differences.

I’m this video you will also learn that much of it is coded in assembly! The developers are prioritizing performance to the extreme, which has its benefits, but I would expect assembly to open up the code to a whole world of vulnerabilities.

We do have an equal temperament scale! That’s what most music uses. What we don’t have is just intonation which uses simple ratios/intervals for the notes. It’s just intonation without the beating, equal temperament, which was listen to all day has the beating but we’re adjusted to it.

I read it all. There are no shockers in the boxes. It's all explained ahead of time and by the time the contents of the boxes are revealed, you'll wish you didn't read all of that.

Thank you I appreciate your opinion. I do think that we are reacting to external stimuli even though our ego is uncomfortable with being deterministic in any way, which is the free will point that you hit on. I think that is likely to be the point that keeps the argument going as it’s not a settled debate absent any AI, which we clearly all see as either deterministic or semi-random when the temperature gets turned up.

As far as the argument about being loaded in memory, if there’s any consciousness in AI, it’s obviously in different form than a biological consciousness. We’d have to agree that consciousness does not require a body to get past this.

The OP said: nozzlegear 19 hours ago [–]

Yes, LLMs don't think on their own, for one; they think when you invoke them.

My rebuttal is that people only think when invoked just the same and can enter states where there is no consciousness just the same. The OP has already accepted that LLMs think, but it seems that you are arguing they do not? This car business is confusing and the LLMs not checking themselves is also wrong, there’s even a benchmark for this https://correctbench.github.io/

exactly! so the arguments against the AI not prompting itself is not a refutation just as it would not be for a person.

Agreed, he’s saying if he’s working with a black box he doesn’t understand what it’s doing. Which is of course true. If the black box were a chinese person instead of a tool and he was writing down what they said, would he say the chinese person in the black box does not understand why they told him to write something? No, of course not.

Today we know the program to do this is always going to be inscrutable to Searle. His role is no different from someone using Claude to write an email.

Are you aware of each of your dreams from last week or last night even?

1) Many people claim to have no internal monologue.

2) We are prompted (invoked) by our environment continuously.

3) If you go unconscious due to fainting or drugs you too will stop thinking.

not really, it’s ultimately formed in electromagnetic fields.

Or it feels just like talking to my grandpa.

AI does not react to endorphins and other hormones so we know that our minds and bodies are influenced by other forces and in other manners. An AI won’t be frightened or angry or aroused, but when those things happen to us they are usually called ‘subconscious’ reactions. Should we require the subconscious to be part of consciousness?

On top of that, neuroscientists find that your mind backfills the conscious reasoning for your reactions after they happen. This is known as our consciousness being the left brain ‘interpreter’ https://en.wikipedia.org/wiki/Left-brain_interpreter

Reading that, an LLM can serve as a ‘left brain interpreter’ as that’s exactly what they are designed to do - come up with reasonable explanations and fill in the blanks based in the data provided.

Right, the export controls are only forcing Chinese AI to innovate, build their own fabs, and make training and inference more efficient. The end game of this will be NVIDIA chips won’t be wanted because you can get a $50 chinese chip running a ternary model that is competitive with claude in English and is much better in Mandarin.

We have GPU costs, power costs, and how many token/s models can generate on those GPUs. It’s possible to figure out the marginal cost based on this. The current estimate is about $0.40 per million tokens for gpt4 equivalent model. Sonnet 4 is $15 per million tokens, so they are charging high margins on inference. The issue is how large of a margin is needed to recover their costs before the GPUs age out, and how high of a margin can be charged before it’s not economically viable.

https://www.gpunex.com/blog/ai-inference-economics-2026/

Coffeezilla bought one of these thinking they’d never be delivered about a week before they announced they would be shipping soon. He wanted to do an exposé on the delays and thought Trump would never release the phone He will now end up with a crappy phone and his personal info exposed

Right, you can rent from a v100 from llama cloud for $0.79/hr. An h100 is $3.99 /hr.

$48000 is equal to 12000 hours of renting an h100, which is about as long as you’d spend at your job for 6 years!