This post is fodder for the neo-conservative (or neo-luddite) left. The neo-conservative left is a political movement by the elite and privileged class of people who are technically literate. Their secret/subconscious ideology is for the status quo to remain. The reason is simple, they never stood for progresss. They actually always wanted to conserve but they liked parroting "progressive" talking points when it helped signal it socially.

This author wants consent from companies deploying LLM. What can one say about this lol..

The AI companies are run by billionaires and now trillionaires that publicly brag about getting rid of artists and creating a permanent underclass. The only people actually profiting off of this are either stakeholders in those companies and/or command extremely high compensation from those companies. Those are the elite and privileged, not the average person who is against this.

yeah? give adequate accurate information to people and let them decide what to do with the information.

if someone told you, you had a .01% chance of getting a disease for example, aren't you better off with that information? even if it is noisy?

> I don't really get what this means. A false positive causes issues inherently - you don't know if it's right or wrong. It's noise which is bad for care, and it's anxiety-inducing for patients which is also bad.

No its not. This is extremely paternalistic. Humans know how to understand noise and statistics. You don't get to decide that for me. I want more lives saved with more information.

It’s not paternalistic. It’s reality. Most people are not equipped to vet all the medical information that comes their way. That’s why we should talk to our doctors about results.

I don’t think my mechanic is being paternalistic when he talks through my car and what is/isn’t important. I like that helps me prioritize things. Why is this any different? In the end a person can tell a doctor “I don’t care run the test” or whatever so what’s the big deal? You can still do what you want. Get that biopsy if you need the peace of mind.

I agree but do you want your mechanic to hide information from you for any reason? For instance, that information can have false positives.

Like, your mechanic hides the fact that your engine may be broken but he's just 70% sure of it. Since he's not 99% sure, he hides it from you. Do you think its a good thing to do?

??? how is more information worse than less? if you know the extent of false positives -- use it! its better than having no information.

the question should be: does cost of obtaining extra information pay off in lives saved. i would say yeah obviously.

AI will be the most important thing to happen to art. We have tolerated low quality art for far too long because we pretend mechanical dexterity is what makes art, art. Art is not valuable because of this. AI removes the part I never cared about. I never cared that a guitarist can physically play 1500 notes a second.

You sound like people who claim AI is freeing because it has removed the gatekeeping of programming when programming was never gatekept, all you needed to do was dedicate time to learn.

Your premise doesn't even make that much sense given that Knockin' on Heaven's Door uses only the three chords(G, C, Am) and they are some of the most common chords in music.

> We have tolerated low quality art for far too long because we pretend mechanical dexterity is what makes art, art.

Who is “we” in this sentence? A tiny subsection of metalheads? The vast majority of people aren’t seeking out the most mechanically dexterous musicians to listen to.

working with a non deterministic tool requires taste and judgement. if you want full determinism AI is not for you but it has a market for other people?

It’s not going to be instant. Just like we have housing shortages or unemployment.

The correct marketing and product strategy is to not stick AI in everything. It’s to allow AI to access them. But this is a hard concept to grasp and tough to give up territory.

A good story here is notion: I don’t think they (only) stuck AI features. They made it possible for me to use it from AI. This is meaningfully different because it enables * composability *.

I record my notes in Notion using Apple Watch and summarise them or use them through Claude account which has a plugin to Notion.

Now think about it: employees in notion wont think of this as an amazing feature because it is utterly simple to implement. There’s no limelight or anything. If they had made some fancy AI integration within notion to autocomplete or whatever, the optics are better internally. But outside it is lukewarm to bad.

I wish more companies enable composability instead of bespoke AI integration within their application.

and he said its cool

No models are subsidised -- there are lots of third party hosting services that will still run at breakeven/profit. (except Deepseek after discount)

> No models are subsidised

We have no proof in either direction, it's not like we had access to their financial numbers in details.

And the pricing itself muddies the water, as input tokens that are already in the KV cache are practically free for the provider, whereas other tokens are expensive. So they could still make money overall thanks to people having multi-turn conversation (and as such, paying multiple times for the same token), but lose money on actual compute done.

> there are lots of third party hosting services that will still run at breakeven/profit.

How can you be sure that they are making profit directly from token price, and are not billing at marginal cost (i.e. electricity price, without counting the cost of the GPUs) and aiming to make a profit later on from the valuable training data that they are collecting in the process?

> as input tokens that are already in the KV cache are practically free for the provider,

not at today's RAM prices.

RAM price don't change anything. You can't fit an infinity of tokens in the KV cache, but the ones that are in there when a user request them are still practically free.

> How can you be sure

You are free to believe that they are doing all this. Or you can simply believe the intuition that models are getting cheaper by the day. I can run Gemma 4 31B from my laptop today.

> Or you can simply believe the intuition

Sure, you can believe you intuition as much as you want, but telling strangers over the internet that they are wrong because “I trust my intuition” is… awkward.

At some point it does come to intuition. Even if the companies IPO and share their financials, you can always argue that they might be lying.

Again, there's a difference between relying on intuition in your life (which we all do, lacking perfect information that would allow us to avoid relying on it), and telling someone they are wrong because your intuition says so.