runlaszlorun

Unless you get your eyes open to Intuitionist Math and then you realize math isn't "true".

Then again... where in the trillion or so parameters of any LLM is The Law of the Excluded Middle that classical math requires to be "true".

Even more comical is that there are certainly embeddings in there _about_ an excluded middle. With thousands of dimensions and billions of values in each one.

Lord help us all... Lol

I havent even RTFA to be fair but I like how often Bayesian heuristics like this turn out to be... Useful... If even not provably "true".

If that buggy Java ridden UI thinks it's "enriching the fascist rich", somebody gonna be pissed fr

Lol

The funny thing is that no one realizes that palantir is shitty java that still need Java 1.1 for AWT, etc

This should scare you immensely more...

#rewrite_it_in_ada

Informative, thanks for sharing.

Oh that's very interesting. Unfortunately for all of us many countries are revisiting nuclear ambitions it seems, but I can see how that makes sense from the Japanese perspective, given an environment with more aggression in general and a lot much reliable US as an ally.

This might be what you're saying, and it's been a long time since I read Rogers' original work, but IIRC Rogers was following the standard deviations of the normal distribution but wasn't without merit in finding common characteristics within said group.

But def incorrect as written in the article.

Good on Switzerland.

It's easy to bash governments but, given that LLMs process basically all of human recorded knowledge and run on top of the collective efforts of 60 years of Moore's Law, we should look at them as a collective asset and need more good actors creating and improving them.

I can't picture any path forward that doesn't involve more customized versions with more agency in shaping them.

LLMs as we have them now are like mainframes compared to the PCs to follow but even more so as they have their own worldview and biases.

> A bubble bursts when debt obligations exceed the ability to maintain that debt for multiple parties simultaneously, which results in a collapsing debt spiral.

Is this true? Not challenging this but hadn't heard bubbles tied to debt instead of market prices on equity or similar valuations.

Interesting point. Would you say the US chip sanctions have given even more incentive for China?

> They can amount to small scale terrorism and that is it

Or a large scale insurgency. Something millions of military veterans have (unfortunately) witnessed first hand.

Agreed. And don't get me started on the whole Thucydides Trap, whose primary trap is that assuming war is inevitable based on conflicts from centuries ago will, in fact, become a stupidly self-fulfilling prophecy.

I think it would be interested if only those who had completed military service could vote to go to war since I think you'd def see less of them. That said, I don't see this actually happening.

I agree whole heartedly with your main point but I'd say militaries are more likely to overreact to the last war. "Generals are always fighting the last war" is a quote I heard from long ago and it seems to match up. In Desert Storm it was clear that the lessons of not getting entangled a la Vietnam were clearly in mind. When I went in to Bosnia it was clear Somalia was in mind as our RoE basically said if we were being being fired up on by someone using civilians as a shield to "aim carefully", and I keep hearing mention that the US is still trying to shift away from GWOT even though it's been 10 years since that started slowing down.

You're completely ignoring the vast amounts thrown into a marketing funnel in an attempt to acquire customers for their usually meh products.

And engineering teams usually scale up with revenue as well.

I feel like your numbers are the myth that gets told not what actually often happens

I love your page. Being an authentic human and actually having some personality seem to be like secret weapons these days.

Would concur 100% on "blind leading the blind" here.

I'm not so sure that loading "more than a few thousand rows" is as bad as it used to be.

I did some quick benchmarks a couple years back. It's been a while but I want to say that Chrome was drawing 10k rows of a decent size each (10 columns of real world data and about 500b/row iirc) in about 300ms on a 10 year old MBP.

I'll do a little benchmarking later today if I get a chance.

It's funny you mention 500kb/s. I have a painfully slow connection at home currently (long story). It's exactly in the 500kb/s to 800kb/s range and "dial up speed" is exactly how I describe it.

Is Elon still super sensitive about ADS-B? Asking for a friend... who lives next to Austin Bergstrom airport. And has 6 rtl-sdr's on his desk begging for mission...

Ah yes... Move face and break things. Well Facebook didn't overpromise on that one...

I'd second the screenshot recommendation.

But def looks very interesting and i like the simple minimal design on mobile. Logo's great too. Congrats on getting it out there!

Completely OT and apologies if rude, but you're Gary Numan the musician?

I love finding out when celebrities have talents elsewhere. And Wikipedia says you've had quite a bit of aviation experience as well.

Kinda makes my morning... lol

p.s. The inner city cynical kid in me is now required to throw in that I found Django disappointing even though every else seems to love it. Ok... identity restored... can now go back to shaking my fist at the sky as per usual...

> The French had a uniquely high Gini coefficient before the Revolution.

I feel like this one line captures the elephant in the room that the author is trying hard to convince himself isn't there...

That's pretty hilarious I need to find that in a poster for a Marine buddy who's a Toyota mechanic.

I'd use the analogy of current dynamic language VMs being written C. You're not modifying structs at runtime or creating C functions as a user of the dynamic language. But the creators of the VM are providing as thin a layer as possible on top of that to give the runtime behavior it's dynamism.

There's an often repeated line that WebAssembly isn't really "assembly" and it's not "web". There's a lot of truth to that. I dove in hoping for a cross platform, web enabled assembly. It looks a lot like an assembly but I find it's relatively few restrictions like lack of a true jump command rippling out more than one would have expected. It's also sortuva stack machine and sortuva register machine at the same time.

It does share a lot with the internal VMs in many places like the JVM, Python's VM, .net, etc.

I love it when two conspiracies start conflicting.

> best recent listen for me was Sean Carroll Mindscape on the emergence of time

Interesting tip, thx for passing along.

It should be added that the company where the blog is hosted makes MCP tools. So a post from a company making MCP tools that MCP is the new Messiah isn't exactly earth shaking.

Sadly this was also my reaction.