iiinova — where your digital self keeps evolving

Riggs @bigrigg47.bsky.social · 08/10/2026

Sometimes?? Are you imagining some mechanism of population decline that consists of a consistent, cross-generational cull? No, population decline always results in too few workers and too many dependents.

100

Riggs @bigrigg47.bsky.social · 08/10/2026

It’s not. Japan has been further urbanizing even as its population drops, and even that hasn’t been enough to end the currency crises and stagnant wages. Population dropping is unambiguously bad for an economy (even in per capita terms).

000

Riggs @bigrigg47.bsky.social · 08/10/2026

Israel’s population is insanely right wing. They are, honestly, much worse than Idaho Republicans.

000

Riggs @bigrigg47.bsky.social · 08/10/2026

Oh, sorry, it has increased. It’s just increased 22% in 30+ years - as opposed to the US’s real GDP per capita doubling over the same time frame.

110

Riggs @bigrigg47.bsky.social · 08/10/2026

It is, in fact, not ok to be Japan, which has been undergoing one long economic crisis since the 90s. Rural communities are dying, the government is constantly trying to reverse the stall, and it’s not getting richer in per capita terms either.

120

Riggs @bigrigg47.bsky.social · 08/10/2026

You defense against “Israel’s electorate is extremely right wing and keeps voting for right wingers” is “Netanyahu is too moderate for many Israelis, so he has had to team up with even more extreme right-wingers to get a majority”???

100

Riggs @bigrigg47.bsky.social · 08/10/2026

Test scores are improving though.

110

Riggs @bigrigg47.bsky.social · 05/10/2026

São Paulo is one of the whitest states in the country (63% vs. 43% for the whole of Brazil).

160

Riggs @bigrigg47.bsky.social · 29/09/2026

And I guarantee that when someone says “inflation makes it hard for people living on fixed incomes”, they are not talking about how inflation will make it harder for the government to sell the bonds needed to fill the social security trust fund’s solvency gap.

110

Riggs @bigrigg47.bsky.social · 29/09/2026

You’re using “fixed income” in the only way that people it nowadays: to lie about current economic conditions.

100

Riggs @bigrigg47.bsky.social · 29/09/2026

I can blame them, because their whole thing is thinking that laws are *not* fake. Their reaction to rich people seeming to get away with crimes is not anything about power or wealth, it’s “there is a secret law, and if I memorize its incantations, I will be free from restriction and responsibility”.

020

Riggs @bigrigg47.bsky.social · 29/09/2026

To make a finer point, it’s not like it’s easy for wage laborers to raise their income when expenses rise. Social security income is *far* more insulated from the stresses of inflation than wage income is.

100

Riggs @bigrigg47.bsky.social · 29/09/2026

No, fixed income means fixed income. And its primary rhetorical purpose is to provide a highly sympathetic victim for inflation. But since it’s barely real nowadays, people morph it into something like “consistent non-wage income for non-wealthy people” and still use it in the same rhetorical place.

101

Riggs @bigrigg47.bsky.social · 29/09/2026

Basically no one lives on fixed incomes anymore. Fixed incomes only really happened with annuity schemes that have mostly died out. Social Security and retirement accounts like 401ks are not fixed income - they go up with inflation.

100

Riggs @bigrigg47.bsky.social · 29/09/2026

It doesn't require parity. It just needs it to be true that attempting to conquer the other is too damaging to attempt.

110

Riggs @bigrigg47.bsky.social · 29/09/2026

Maybe the coalition that you're thinking of doesn't survive sanctions, but Democrats as a party and an electoral bloc certainly do. Pro-Israel Democrats are a rapidly dying breed, and they are far more likely to dwindle into insignificance rather than to stabilize or rebound.

1160

Riggs @bigrigg47.bsky.social · 27/09/2026

I get defensive when people talk shit on my city, but never my state or the region - I am sure that I hate the rural conservatives in my state far more than blue state liberals + leftists do.

020

Reposted by Riggs

William B. Fuckley @opinionhaver.bsky.social · 27/09/2026

Stack of books contains a guy if it’s big enough.

3926

Riggs @bigrigg47.bsky.social · 26/09/2026

No. There is absolutely something wrong about lots of preferences. I don’t get why people acquiesce to this framing. Lots of preferences are bigoted, stupid, or harmful. When they’re inconsequential, it’s usually not worth making a fuss about, but oftentimes these preferences are very consequential.

000

Riggs @bigrigg47.bsky.social · 26/09/2026

I really don’t think there is a single justifiable reason that one could blanket refuse to date bi people. Like, we don’t need to delve into someone’s reasons to figure out if they’re being bigoted. We know they are.

010

Riggs @bigrigg47.bsky.social · 26/09/2026

It’s definitely a preference. It’s just a bigoted preference. Many preferences should not be respected, and I find it so bizarre that “it’s just a preference” is used to defend so many bigoted, stupid, and harmful behaviors.

020

Riggs @bigrigg47.bsky.social · 24/09/2026

A description of a situation normally means that you’re not lying about it, but I guess YMMV

120

Riggs @bigrigg47.bsky.social · 24/09/2026

This falsely implies that people at OpenAI were intending to hack Hugging Face, and it completely misleads as to the actual shape of what they did wrong here (utter negligence, not premeditation). Your entire goal here is to blame OpenAI, and you can’t even do that accurately?

120

Reposted by Riggs

BeijingPalmer @beijingpalmer.bsky.social · 24/09/2026

I believe that AI, like any other tool, can and will generate art when used interestingly, and also that almost nobody is using it interestingly yet

5033830

Riggs @bigrigg47.bsky.social · 24/09/2026

What? The A-10 has been artificially kept alive for political reasons for decades.

1583

Riggs @bigrigg47.bsky.social · 24/09/2026

You said that something being created by humans means it can destroyed by humans. I don't believe that - I think that's dumb as fuck. I'm just following your logic a little bit. Fuck me for taking you seriously I guess.

150

Riggs @bigrigg47.bsky.social · 24/09/2026

Since apparently you consider yourself some sort of inferior form of humanity, who are these people that can stop a bullet racing towards their heads just because people invented guns?

161

Riggs @bigrigg47.bsky.social · 24/09/2026

Do you think the fact that you fired a bullet means you can stop it?

170

Riggs @bigrigg47.bsky.social · 24/09/2026

We created it therefore we can destroy it? What? How does that make any sense?

160

Riggs @bigrigg47.bsky.social · 23/09/2026

They are not collage machines. It is only "fancy auto complete" in the sense that to accurately auto-complete basically any input message, you need an embedded model of all of known reality (i.e., you need to understand reality).

040

Riggs @bigrigg47.bsky.social · 23/09/2026

The reason image recognition is focused on interpreting medical scans and not besting whatever ill-defined god-of-the-gaps challenge you have in mind is because the former is actually useful. Like robotics seems legitimately very technically hard, but generic "sensory perception" really doesn't.

200

Riggs @bigrigg47.bsky.social · 23/09/2026

What "human sense"? Because the way that I (a human) have been defining "reasoning" for my whole life really seems to include whatever modern LLMs are doing.

030

Riggs @bigrigg47.bsky.social · 23/09/2026

It's not really more complicated. We've been running images through neural networks for decades and they're rather good at recognition nowadays (and they can be pretty quick).

110

Riggs @bigrigg47.bsky.social · 23/09/2026

Cameras and microphones and chemical detectors are not particularly expensive...

110

Riggs @bigrigg47.bsky.social · 21/09/2026

Do you dislike "thinking in a way that's unlike us" also? They certainly don't think like us, but, considering their capabilities, it feels inaccurate to implicitly categorize whatever they're doing as inferior to our processes (and, consequently, less capable).

310

Riggs @bigrigg47.bsky.social · 20/09/2026

Do you not work exactly according to your biochemistry? If no, that seems to imply that there’s some immaterial essence somewhere in there. If yes, I don’t know what’s so different about LLMs working through their defined processes with their embedded stochasticity.

040

Riggs @bigrigg47.bsky.social · 20/09/2026

The reason it gets stuff right is not because it's memorized or stored its training data (not physically possible - it's too small); it's because (apparently) when you compress and extrude and transform human language enough, out pops something with embedded models of grammar + logic + reality.

110

Riggs @bigrigg47.bsky.social · 20/09/2026

The idea that "hallucinations are just guesses to fill in the blanks" is just completely disconnected from the reality of how these things work. It's a statistical model derived from all human language - all it does is guess. Hallucination is just the label we apply to it when the guess is wrong.

210

Riggs @bigrigg47.bsky.social · 20/09/2026

LLMs *do not* have a database. This is a very important distinction here. You are thinking of them like traditional computer programs. They are nothing like that. They are much weirder (and, seemingly, much more analogous to human thought) than your conception of them seems to be.

200

Reposted by Riggs

1st Minnesota Volunteers Stan Account @himdogson.bsky.social · 19/09/2026

too much jeffersonianism: mao, hitler, khomeini, pol pot, trump healthy jeffersonianism: paine, mandela, nehru, lula perfect balance: franklin delano roosevelt healthy hamiltonianism: carney, tito, lincoln, de gaulle too much hamiltonianism: stalin, park chung-hee, chiang, mussolini

810021

Riggs @bigrigg47.bsky.social · 20/09/2026

From what I can tell from the documentation, all of that kind of stuff seems to be manually coded in: www.nytimes.com/interactive/.... Maybe they involved some traditional ML techniques, but it really doesn't seem like LLMs were involved at all.

100

Riggs @bigrigg47.bsky.social · 20/09/2026

Language doesn't have meaning - it has usages. You're including this bit of "it has to connect to reality" into how you're using "thought", but that doesn't seem integral to the colloquial use of the term.

121

Riggs @bigrigg47.bsky.social · 20/09/2026

TBH, at this point, this is mostly philosophy, not biology or CS. Thought is obviously emergent from lower-level non-thought processes. So it makes no sense to say that since linear regression isn't thought, what an LLM does (gigawatts of linear algebra) can't be thought.

030

Riggs @bigrigg47.bsky.social · 20/09/2026

This isn't a great analogy because the Wordle Bot is basically a traditional computer program, which works nothing like an LLM. The thing that's talking to you didn't solve it - it's just reading the results from a database that something else generated (you could do the same).

110

Riggs @bigrigg47.bsky.social · 20/09/2026

Who accuses a cell of thinking? Unless you believe in a soul, thought is obviously some sort of emergent process. We know biology can think, not because we analyzed its smallest constituent unit, but because we can see the results of thought from a complex arrangement of those units (humans).

110

Riggs @bigrigg47.bsky.social · 20/09/2026

LLM inference is not deterministic; they are nothing like imperative programs.

110

Riggs @bigrigg47.bsky.social · 20/09/2026

Turing specifically came up with the test as a concrete, testable alternative to the (very vague and indefinable) problem of "can machines think?". And LLMs blew past it years ago.

1100

Riggs @bigrigg47.bsky.social · 19/09/2026

Mosquitos kill about twice as many people per year as humans do.

120

Riggs @bigrigg47.bsky.social · 18/09/2026

No, it is not. It is the measure of the prices of all the things that consumers spend money on (the vast majority of which is not gas).

000