Gunson

Zero-JS Hypermedia Browser

avatar
Gunson
gunson@primal.net
npub1pn9x...2xn0
Low status fiat heretic. Often wrong. 2 + 2 = 4

Notes (14)

Seeing lots of Bitcoin price overconfidence "Stock to flow says we're about to hit $400k!" "There will be no more winters!" "Banana zone something something about to rip!"
2025-06-11 06:48:49 from 1 relay(s) View Thread →
It doesn't matter how many variables you have, you can't improve the modelling of a quadratic function with more than 3. You also don't improve your model with more data points past some amount. This analogy describes some of my skepticism of the claim that we will create super intelligence by merely adding more data, more parameters, and more compute to LLMs. Yes we will continue to see improvements in the ability to model human knowledge and reasoning, and yes it will be amplified by machine speed, but I don't think we somehow unlock a higher level of intelligence. We will plateau at some point and then the focus will shift more to applications of this still impressive but sub-human intelligence.
2025-06-10 06:58:40 from 1 relay(s) View Thread →
Why isn't Bitcoin $0?
2025-06-02 06:42:05 from 1 relay(s) View Thread →
Need to do more fiat shaming
2025-05-31 07:28:12 from 1 relay(s) View Thread →
New global warming propaganda: Glaciers falling on tiny Swiss villages
2025-05-31 07:02:26 from 1 relay(s) View Thread →
AI is really good at giving me an answer that a reasonably smart expert would. That's pretty useful! But I'm still slightly skeptical about how it exceeds that without just relying on crunching more data - will it genuinely understand things at a higher level than best human level?
2025-05-21 18:02:16 from 1 relay(s) View Thread →
Kid number 2 arriving in 2025 ♂️
2025-05-20 16:03:33 from 1 relay(s) View Thread →
Impressed with Mullvad's marketing. Their brand and messaging is all over London making a strong case for privacy 👏
2025-05-16 09:50:20 from 1 relay(s) View Thread →
Whenever I ask an LLM about something I actually know a lot about it's quite obvious how much it bullshits and avoids directly answering a question, or giving an actual reference.
2025-05-13 08:11:16 from 1 relay(s) View Thread →
Some of the smartest people in the world are warning of AI taking over governments and replacing everyone's jobs in ~3 years. Haven't seen it mentioned once in mainstream headlines. But we're treated to stories about Mexican gang members working in call centers, and how today is the hottest day of spring 2025.
2025-05-01 12:49:36 from 1 relay(s) View Thread →
Left wing socialist parties are always described as "centre-left", and parties that want to steal less of your money are always "right wing" or even "far right". Fascinating.
2025-04-29 09:24:24 from 1 relay(s) View Thread →
As I read about the Napoleonic wars it strikes me that calling one of the belligerents the "aggressor" is a peculiar modern term. It was never done, since it was clear that all were seeking power in different ways due to different interests. While Napoleon was bringing proto-democracy to regions, he was also an autocrat, and was an economic rival to England. The English funded other countries to confront France, yet it was France who was imposing the Continental System and taxes on their allies which caused them to switch allegiance and declare war. These wars would usually end with one of the belligerents needing to make major territorial, military and economic concessions. The modern framing as one of a victim (Ukraine) and an aggressor (Russia) seems like a modern propaganda tactic designed to prolong a war and avoid the historically typical approach of ending it with a peace negotiation where the stronger side militarily has more leverage.
2025-04-24 13:39:26 from 1 relay(s) View Thread →