root_axis

A big difference is the existential levels of debt.

> Then why have they been lagging behind OpenAI and Anthropic

Because they're not desperate. Slow and steady wins the race, at this rate all Google has to do is wait for OpenAI and Anthropic to exhaust themselves on aggressive training, then they can casually amble along right past them.

More productivity might translate into more revenue, but more productivity doesn't necessarily create more wealth.

For example, if I'm a landlord that uses AI to more intelligently price rental increases to maximize profit without attrition, I have created more revenue but not more wealth.

I don't think it guarantees job security, I think the future is an adaptation to the prodslop paradigm where the answer to vibe-coded monstrosities is more LLMs.

Ultimately, the industry will be forced to adapt to the extinction of the junior to senior pipeline, but investment in more tokens is going to be more attractive than paying specialist engineer salaries, especially since LLMs probably have a few more generations of qualitative improvements before diminishing returns truly set in.

> C/Rust/Go wrap assembly. TypeScript wraps JS. Java/Python wrap C.

LLMs rely on the semantic richness of these languages in order to capture the the gradient of human intent from their training.

There is no "problem".

The architecture and the intent don't matter to business folks, it never has, and with LLMs it matters less than ever.

Vibe code into production is satisfactory regardless of architectural understanding or intent. If there are issues, just have the LLM spin up some agents to play whackamole until the issues are pushed beyond visibility.

Ultimately, the idea that we can hang onto fleeting engineering disciplines misunderstands where the industry is going, regardless of any assessment of the LLMs capabilities.

Do you restart the entire sentence in your head every time you add a new word?

What is an LLM going to do with food, clothes, diapers etc? The idea doesn't really make sense.

I think you could make a corresponding argument about the unbounded complexities of an abstract domain like mathematics. Math also has a notorious reputation for difficulty. We don't even have a way to know how much there is to know about math.

I'm not sure I see one is clearly more difficult than the other.

This framing overlooks an unstated caveat - i.e. people that work for an LLM company have an incentive to minimize human contribution as much as possible in their narratives.

I don't think that's a given - not sure how you'd even quantify that comparison.

That's of course absurd. The human brain doesn't represent information as discrete tokens, nor does it form sentences autoregressively.

Doesn't really seem possible. The two party system is so deeply wound into the structure of the u.s. political machinery that a new party doesn't really stand a chance.

The consistent through-line of Greenwald's journalism from the Bush era through to the Trump era is an anti-u.s pro Russia narrative.

Not likely, both political parties hate him.

> Just as a simple example look at the golden ratio in living organisms

What is this example meant to illustrate?

What is there to figure out? Elections have consequences and the country is learning that the hard way... again.

Do I blame the people with all the power or the people with none? Hmm.

> We have layers of management, which we actually have found helpful

Interesting discovery.

It's still spam when a human does it.

You're thinking about this all wrong. Just step into this cranial MRI scanning booth to livestream the structure and composition of your skull straight to the cloud, then we'll fire up 10000 agents to find the perfect cap for you.

> Make it so the model can't misbehave

Not possible. They can chase the models with whack a mole tuning for obvious stuff, but there's always a way to extract what you want from the model.

> Why would autocomplete know the moral difference between breaking out of its working dir and hacking a package manager?

I don't think it's even a question of distinguishing "moral difference", it just comes down to the "stochastic parrot" behavior that people hate to acknowledge. Yes, at these absurd scales the LLM can maintain impressive levels of coherence, but at the end of the day, spinning up 10000 agents is just running a tree of 10000 prompts in parallel, some of them are just gonna do wacky shit, with the harnesses acting as homeostasis for tasks spiraling into nonsense.

Who could possibly hold them accountable?

Even if legality did matter, the president has criminal immunity for official acts.

Even vanilla minecraft is a far superior experience to NMS. Minecraft's approach to procedural generation poduces more meaningful exploration because its terrain, resources, structures, and building system interact far more deeply with the gameplay than NMS linear number-go-up grind.

Of course, introducing mods makes Minecraft the most compelling entry in the genre by a huge margin, but not necessarily a fair comparison since the combined human effort involved in the minecraft mod ecosystem vastly eclipses what the NMS team could do on their own.

Poor reasoning. You could apply the same logic to any improvement in any field of machine learning.

The evidence is the history of the sun rising since time immemorial, as well as the science of physics and cosmology that models the sun's motion with respect to the earth.

This is olympic level mental gymnastics to justify believing things without evidence. The double negative with the word evidence in scare quotes is chef's kiss.

It's incredible to me that every single time there's a new model people scream "singularity" from the rooftops and every time they are wrong.

This is an impressive result, but there is absolutely zero evidence of "the singularity".