eaglelamp

Call it what you want, but if the companies who received the orders are able to treat them as assets/income then the overall macro effect will be the same as with debt: leverage. If funding begins to dry up it may not be Anthropic or Google facing a liquidity crunch, but someone will be.

The USSR used American racial discrimination in its propaganda as well. Does that mean that civil rights activists in the US were internal enemies or that effective propaganda highlights legitimate grievances?

Perhaps the current elite should focus on reasonable ways to address mass discontent rather than justify intensified oppression.

What do you mean by "fundamental truth"? If LLMs produce a collection of proved facts this doesn't sound like what I would call fundamental. It may be the case that there is no such thing as "fundamental truth", that the world is just a blizzard of facts, but in that case we would need to give up on using it as a guiding star.

Arguments for LLM consciousness are a Trojan horse for strengthening the rights of corporations. How could a model trained, controlled, and operated by a private corporation be anything except for an extension of that same corporation?

This has ramifications for assessing their consciousness as well. Conscious experience does not pause as you wait for input from a puppet master.

The first section of this post responding to "Meta, Google, and Microsoft are dying" is completely misrepresenting Zitron's arguments in the linked video. Zitron is making essentially an enshitification argument concerning Google and Meta. He's not talking about whether they will be able to squeeze out more revenue in the short run, but how they are willing to abuse their users to ensure that they do. Citing increased revenue for these companies is not arguing against Zitron's point.

Luu also fixates on Zitron mentioning Prabhakar Raghavan, but then proceeds to agree with Zitron's core point that Google has intentionally degraded it's search product to maximize revenue. Maybe Raghavan is not solely responsible, but it seems fair to hold him accountable for trends that accelerated under his leadership.

My cognition is outmatched by predicting the impact of throwing a brick over my neighbor's fence. I have no idea if it will land harmlessly in a patch of grass or fracture her skull. Once I've thrown the brick, even if I see my neighbor enter her yard, my reactions are too slow to save her.

I'm not the wisest man, but I'm wise enough not to throw the brick and see what happens.

Similarly OAI should have the wisdom to see that deploying a hazardous swarm of agents with access to the public internet could result in harms, and that those harms would manifest quicker than humans can react, but they unleashed the swarm anyway.

I'm not.

As it stands LLMs are not intelligent, they have no agency, they only produce output in response to input. Ultimately this input comes from a human who is an intelligent agent and should be held responsible for the consequences.

Humanity has created and tamed many dangerous tools. Creating a fantasy world where LLMs are super intelligent and beyond the control of any mere mortal isn't going to help us build the norms that minimize their harms.

Of course there is more to be learned by exposing the entire world to your dangerous creation, that doesn't justify doing it. I'm sure we could learn a ton about infectious diseases by designing new ones and unleashing them on the world, but there are very good reasons why we don't.

Most of the benefits could have been gained from a network isolated from the internet. OAI could have deployed servers to exploit and methods for inter-agent communication on such a network easily. They could have even worked with partners to deploy cloned versions of their infrastructure in this sand-boxed environment.

The only problems with an isolated network approach are: it takes some amount of effort, and it doesn't create another "AI apocalypse" news cycle.

From Wikipedia:

>After graduation, she worked on Wall Street, first at JPMorgan Chase and then Lehman Brothers. She later joined the United States Foreign Service.

Looks pretty wall street to me.

Also you regarding Claude usage limits:

> Before the doomers come in, you get $200 in API credits every month for claude -p usage. Usage counts against those API credits.

So which is it $300/day is trivial to consume or $200/month is a completely reasonable limit, it can't be both.

Inference might be unreasonable for a royalty agreement, but, in assessing damages, it is certainly relevant.

"I made enough copies for everyone" isn't a valid defense for copyright infringement.

You're misinterpreting the quote. Socrates is saying that being able to find a written quotation will replace fully understanding a concept. It's the difference between being able to quote the pythagorean theorem and understanding it well enough to prove it. That's why Socrates says that those who rely on reading will be "hard to get along with" - they will be pedantic without being able to discuss concepts freely.

Likewise with AI the appearance of reasoning without the substance could lead to boring exchanges of plausible slop rather than meaningful discourse.

China is not the only nation that depends on Gulf oil, all of SEA does as well. If the strait remains closed it will destabilize the region and diminish the prestige of the US, and with the US military focused on the strait China will benefit.

Current US interventions should be read as a sign of weakness - an inability to shape events without resorting to naked aggression. Global hegemony can not be maintained through force alone.

No one ever asks how much it costs Facebook or Uber to serve requests because it is irrelevant, they set prices to maximize their profit like any good monopolist. Similarly the future cartel of big providers will charge their captive users whatever they can get away with, not the cost of inference.

The current discourse around "AI", swarms of agents producing mountains of inscrutable spaghetti, is a tell that this is the future the big players are looking for. They want to create a captive market of token tokers who have no hope of untangling the mess they made when tokens were cheap without buying even more at full price.

If they have spare capacity then there is no opportunity cost to selling $100 subscriptions for exactly that reason. If they don’t have spare capacity then, at the margin, they could replace a subscription user with API calls that make them $5000: that’s opportunity cost.

If you own equity in Anthropic you should care about that cost. Maybe you are willing to tolerate it to win market share, but for you to make the most profit you need that cost to shrink.

You’re looking through the wrong end of the telescope. An investor is buying opportunity and it is a real cost to them.

If Anthropic's compute is fully saturated then the Claude code power users do represent an opportunity cost to Anthropic much closer to $5,000 then $500.

Anthropic's models may be similar in parameter size to model's on open router, but none of the others are in the headlines nearly as much (especially recently) so the comparison is extremely flawed.

The argument in this article is like comparing the cost of a Rolex to a random brand of mechanical watch based on gear count.

Anthropic knew they were going to lose this contract to OpenAI, and this is an attempt to salvage publicity from the loss.

This administration is comfortable with blatantly picking winners and OpenAI is better connected with the admin than Anthropic.

> The politicians' initial efforts caused a public outcry though, when they stipulated that all samples should be destroyed on a donor's death.

Looks like you're right.

As an investor in Anthropic which pricing strategy would you support? That's the question you need to ask, not what there current pricing strategy in the win the market phase happens to be.

In my experience the most common use of this data is to build case for firing someone for cause when upper management wants them out. It's rarely used for actual security purposes.

Capitalism is the only economic system that has the privilege of being evaluated outside of the context of the society in which it exists. When socialism is criticized the political system is always, justly, included; so the purges of Stalin and the homophobia of Guevara are all taken into account. Apply the same thinking to capitalism and you have to count a lot more deaths and injustices: the Irish and Bengali famines, world war 1, climate change, etc.

You're misinterpreting the quote. Socrates is saying that being able to find a written quotation will replace fully understanding a concept. It's the difference between being able to quote the pythagorean theorem and understanding it well enough to prove it. That's why Socrates says that those who rely on reading will be "hard to get along with" - they will be pedantic without being able to discuss concepts freely.

All employees solve problems. Developers have benefited from the special techniques they have learned to solve problems. If these techniques are obsolete, or are largely replaced by minding a massive machine, the character of the work, the pay for performing it, and social position of those who perform it will change.

Normal Erlang code has a fixed number of reductions (function calls) before it must yield to a scheduler. Processes also have their own stacks and heaps and run garbage collection independently. The result is that no single process can stop the whole system by monopolizing CPU or managing shared memory.

The Erlang runtime can start a scheduler for every core on a machine and, since processes are independent, concurrency can be achieved by spawning additional processes. Processes communicate by passing messages which are copied from the sender into the mailbox of the receiver.

As an application programmer all of your code will run within a process and passively benefit from these properties. The tradeoff is that concurrency is on by default and single threaded performance can suffer. There are escape hatches to run native code, but it is more painful than writing concurrent code in a single-threaded by default language. The fundamental assumption of Erlang is that it is much more likely that you will need concurrency and fault tolerance than maximum single thread performance.

Higher level abstractions are built on rational foundations, that is the distinction. I may not understand byte code generated by a compiler, but I could research the compiler and understand how it is generated. No matter how much I study a language model I will never understand how it chose to generate any particular output.

I don’t think the author is hiding his economic anxiety behind solipsism. He states plainly he doesn’t like the deskilling of work.

My point is why are your economic motivations valid while his aren’t?

If you dig deep enough isn’t the same thing true of people like yourself? Do you truly believe that the large language models we currently have, not some fantasy AI of the distant future, are emotional and intellectual beings? Or, are you more interested in the short term economic gains of using them? Does this invalidate your beliefs? I don’t think so, most everyday beliefs are related to economic conditions.

How could a practical LLM enthusiast make a non-economic argument in favor of their use? They’re opaque usually secretive jumbles of linear algebra, how could you make a reasonable non-economic argument about something you don’t, and perhaps can’t, reason about?

The primary harm of a bubble is *not* a crash in equity values, it is the misallocation of capital. The worst outcome would be for the misallocation to continue due to the intervention of asset owners with the most to lose who are also in control of the state.

All of the events you listed have had significant economic effects and required massive intervention from the state to buoy asset prices. The longer this continues the more our economy becomes geared to producing "value" for this small, and shrinking, group of owners at the expense of everyone else.

It looks like security/surveillance play more than anything. Scale has strong relationships with the US MIC, the current administration (predating Zuck's rebranding), and gulf states.

Their Wikipedia history section lists accomplishments that align closely with DoD's vision for GenAI. The current admin, and the western political elite generally, are anxious about GenAI developments and social unrest, the pairing of Meta and Scale addresses their anxieties directly.