agnosticmantis

How would these compare to a MacBook Pro M5 in terms of performance and price?

> If AI is good enough that it can boost productivity by 20% then it is good for society in general because the gains will be redistributed (as it has always been)

This redistribution has never been as automatic and inevitable as you seem to assume.

> What better thing can happen to the curious ones amongst us to get an oracle that can answer every question?

Getting paid to answer said questions would be nice. The alternative is you'll have to work 14-hour shifts in a warehouse to be able to pay for ChatGPT 10.0 subscription with ads, but sure, it can answer your math problems and satiate your mathematical curiosity.

Both can be true: insiders bet on a sure thing, outsiders naively gamble, the house gets a cut no matter what.

> Ideas like a linear timeline progressing from A to B (rather than a cyclical one)

Seriously? So we’re either linear Christians or circular Hindus, and nothing else ever existed?

And nobody else believed in linear time?

Likely because Grothendieck was a vegetarian, at least later in his life.

It may not be mainly or solely due to LLM pollution, but rather the fact that every publisher, (social) media company, newspaper, etc. clammed up and started charging (licensing) fees sometime in the last couple of years.

So maybe there's just not much openly available and new content worth training on that wasn't available prior to 2025.

> I lived through the great MTBF vs MTTR (mean-time-between-failure vs. mean-time-to-recovery) reckoning of infrastructure during the transition to cloud and cloud automation.

What's the historical context for this MTBF vs. MTTR reckoning?

Willing to bet they'll lobby for regulatory capture and raise the drawbridge for the little guys.

Now discuss whether a bonobo, a dog, a cat, a mouse, an ant, a bacterium is conscious.

And you’ll find it’s not as clear cut.

My $0.02: Competing against exp(t/2) + exp(t/2) is much much easier than exp(t/2+t/2)=exp(t).

(If anthropic didn't exist, ØpenAI would suck up all the capital and talent in the room. Anthropic's existence has helped divide capital+talent that'd otherwise be gobbled up by the single fastest growing player.)

"You could see a single robotaxi being worth, or providing, about $30,000 of gross profit per year. ... A Tesla is an appreciating asset..."

- Elon Musk during Tesla's Autonomy Day in April 2019.

> Checkpoints are a new primitive that automatically captures agent context as first-class, versioned data in Git. When you commit code generated by an agent, Checkpoints capture the full session alongside the commit: the transcript, prompts, files touched, token usage, tool calls and more.

Whether or not useful for agent collaboration, the data here will be more valuable than gold for doing RL training later on.

Archived (non-paywall) version: https://archive.is/BObYG

One would have to control for all the layoffs, tarrifs, interest rates etc too.

On the other hand, people are working much harder today than 3 years ago (remember people not showing up to work and posting on TikTok about how little work did collecting paychecks from 2 different companies etc?)

Just saying it's very hard to look at a time series and determine an effect size, even though politicians/CEOs like to claim ownership for growths.

I’ve started experimenting with Quarto[0] for scientific/technical publishing on a personal website, and it’s been quite easy to use so far. I especially like that it has builtin support for LaTeX, markdown, code blocks and Jupyter notebooks. Only thing is I wish there were more templates ready to use.

[0] https://quarto.org

He shamelessly advertised companies in his portfolio, whether it was always having a can of coke or candies on the table or stopping at a McDonald's in a documentary.

I refuse to believe that his lifestyle was what was on display.

He lived in the same house for 60 years, sure, but his private jets kept getting upgraded. Good for him, but I find the frugality theater very off putting.

I find it hard to see an upper bound to demand for compute/power.

If models/hardware become 1000X more efficient/cheaper/smaller, people will be running 1,000,000X more jobs with even bigger models on even more intensive tasks, running them for even longer, etc.

We haven't even scratched the surface for video, world models and robotics.

When the markets panicked during the first DeepSeek episode, I just bought more Nvidia for this very reason. NVidia may very well tank sooner or later, but I'm sure it won't be because demand for compute tanked due to more efficient models. A more plausible scenario would be a competitor dethroning NVidia, kinda like Google's recent surge indicated.

This says Gemini 2.5 though.

The Pipe Query Syntax in GoogleSQL implements this elegantly as well:

https://docs.cloud.google.com/bigquery/docs/reference/standa...

> Once AGI is declared by OpenAI, that declaration will now be verified by an independent expert panel.

So OpenAI will declare AGI as soon as ChatGPT is a better AI lawyer than any Microsoft could hire.

This looks very cool, I’m gonna try it later today.

Out of curiosity, why the name Erdos? AFAIK Erdos was neither a statistician, data scientist nor AI researcher.

He sure solved many probability/combinatorics problems and famously had many many collaborators.

"His decision seems to have been based less on his famous scientific thinking and more on a very personal feeling. He wanted to retain his autonomy until the end and to shape his own end."

So you could say it was more system 1 thinking rather than system 2.

I would've expected the opposite given our survival instincts.

Only difference is soon you'll have Mickey Mouse tell you (and your children) to buy same old crap that the influencers were shilling.

Business idea: disrupt all these apps by creating an agent that watches all the brainrot for us so we all get our stolen attentions back.

How about the donor piggy though?

If this becomes commonplace and animals are bred/raised just for their organs, we get into murky ethical territory. (Yes, people already eat the organs for food, that's murky too. But industrial scale organ farming sounds even worse somehow.)

It’s wild that $goog is so undervalued (p/e 27) given Alphabet owns Waymo in addition to everything else, and yet Tesla is so overvalued (p/e 243!!!) despite zero Robotaxis in the near (or far) future and lackluster sales.

Goes to show empty promises and fraudulent showmanship sell better than actual working products that people use.

The ingroup circle of "we're not them (the outgroup), so we're safe" keeps getting smaller and smaller.

And the ads haven’t even showed up in the night sky yet!

It’s going to be much sadder once we start seeing McDonald’s ads wherever and whenever we look up.

Or as Anthropic's Safety team would write: "Trees conspire to take down human race by sabotaging the underground water network."

Unless the whole presentation was generated using sora-gpt-5 or something, this was very underwhelming.

We know for a fact the slides/charts were generated using an LLM, so the hypothesis is not totally unfounded. /s