You’re not going to context hack complexity, every time you summarise something and pass it onto the next agent. You’re losing complexity.

We’re going to get some massive models with incredible context, but AGI requires models to never hallucinate. Even when the complexity requires more context than it has available, which is not feasible

This is the primary paper I reference.

https://arxiv.org/pdf/2507.07505

Vishal Sikka, advisory board member of BMW. Recommended to Stanford by Marvin Minsky, one of his professors were John McCarty. And I must stand corrected, he has a PHD of computer sciences, not Math as I remembered it as.

Varin Sikka is his son, co author of the paper and based on Stanford’s site an undergraduate. https://profiles.stanford.edu/363374

Vishal has an AI based company himself, so there might be some personal reasons for why he’d advocate for using what AIs capable of rather than chasing an impossible (from his perspective) to hit milestone

We have math PHDs with proofs current LLMs can’t become AGI, because they’ll always have a context issue

Just because we’re on the echo chamber of Reddit rejects, doesn’t equal Reddit being minuscule.

Reddit is still living with a huge base of users.

Yep, most people would not understand Machine Learning or Big Data analysis generally being AI, so just market your product by the sub category and you don’t have to deal with the expectancy of GenAI

Honestly it seems like they more or less identified all of the former unidentified users, and it turns out the vast vast majority of which are on unix based systems.

If this goes right down again, I’d imagine the unidentified users to go up - But would be a decent indication for a lot of traffic being scrapers

I honestly think it would do a lot if all sorting algorithms were made open source.

We would quite quickly have mapped out how their algorithms work and which companies are the most predatory

Sounds like they figured out most of the previously unspecified traffic, comes from Linux based systems.

But you’re right about an unknown amount of that traffic is from servers.

It does make some menial tasks simpler, but if people think it will actually replace us then they’re in for a shock.

That’s the sales pitch, we’re in a job market drought right now. Give it a year or two and all the slop will begin breaking, making a new job market boom for developers

Given EU have created laws to require physical knobs for key features, I’d imagine this will be stopped as well.

BMW is the brand that tests for profit initiatives to see if they stick, I recall them being reported to attempt seated heating as a subscription a few years back

A human brain model would never get us to a super computer, which is their intended goal.

Our brains have about one exaflop of processing. The NVidia B300 has 15 petaflops of processing. Corresponding to 0.015 exa, you’d need 66 of the strongest GPUs on the market to match the processing power of one human brain.

That’s before considering context, current estimations would put the human brain at 2.5 petabytes of storage. The former B300 has 288 gigabytes of VRAM. Corresponding to 0.000288 of one human brain, the earlier expectation of 66 landing flat at 0.019 the storage of a human brain.

Can arguments be made not all brain capacity is at VRAM at all times? Most likely. But context windows is the primary bottleneck, not compute

Our brains are absolute monsters, we’d gain more from understanding how we can have this powerful of a machine on so little space. The bigger breakthrough would be brainbased processing of software, but I’ve not heard of anyone going that route - Probably because it’s not feasible

Although simulations agree with you, that LLM powered agentic frameworks are more efficient than human CEOs at optimising a company. They’re also incapable of feelings, making the board of directors fire them in the majority of simulations, for being too cynical

Replacing CEOs with agentic LLMs, is not a benefit for the workers. Far from it, unless you prefer an openly psychopathic CEO

There’s also laws in place if you’re in focus, that requires permission.

Sure I can take a picture of a plaza and if you happen to be on that photo, that’s bad luck. But I can’t zoom in on you garbling down a spinach roll.

For your scenario of going too close and forcing their way to film voyeur stuff, sounds like harassment

I do not agree Microsoft is in that category, but they’re mainly a B2B player today.

A little less than 20% of revenue per earnings '25 was from consumers - The rest for business offerings. I’m curious to hear which selection of companies you consider the main ones. Because Microsoft is still top 5 within Tech when it comes to power.

Google hit as big a market share as they could, so their way to grow the search machine was to increase the market.

How do you increase the market? By pushing the most relevant links further back, so people have to go to page 2, 3 or 4. So it’s unfortunately the best financial decision, short term, to enshittify their main product

It’s mostly all of the generated by LLM websites that’s annoying to shift through

As long as blue ray discs are released, there’ll be webshops selling them

For digital media, pirating increases the number of sales

Why would you make it harder than it is?

libre office and you got an open source Office suite

I believe that’s very regional, if there’s a local hiking club you could try to reach out to them.

There’s often forums dedicated to hiking, which have all sorts of hikes mapped out.

But a map service like mapy.com (formerly .cz), has the official marked routes as part of their mapping data