Andy "Andyfa" Schwarz

Andy "Andyfa" Schwarz

@andyhre.bsky.social

Dorky antitrust economist with an interest in sports economics, athletes' rights, and way too many boardgames.

1.7K followers812 following14.1K posts

Wait, I thought MAGA says Cheesecake Factory is classy. Why is he mad then?

Aaron Rupar@atrupar.com

Trump is BIG MAD about Obama saying the White House now looks like Saddam opened a Cheesecake Factory: "He's not a smart guy. He's not smart at all ... for you to say it's a Cheesecake Factory, that just shows why you were a lousy president"

Andy "Andyfa" Schwarz reposted

If you bought an MLS club right before the league got Lionel Messi, the Apple deal and a domestic World Cup you would have been better off putting that money in an index fund

Tom Lowry@tomlowry.bsky.social

The stock market 📈 vs. team ownership 💪 .. who wins? @levakabas.bsky.social for @sportico.bsky.social ...

This i probably a controversial take, but I find all the extra bells and whistles added to Virgin Quenn vis-a-vis Here I Stand to really weaken the game. Every time I have to interact with a new system in VQ, it tires me out rather than excites me. (more)

If the games had come out in the other order, I would say that someone had removed the bloat from VQ to make HIS a masterpiece. Unless, of course, I have first-learned bias! But I do not think so. I keep trying to like VQ and it keeps not happening. Like fetch.

speaking of the algorithm, I went to read about the awesome Somali guy who rescued a runaway bus filled with kids on the Good Morning America (ABC) website, and got shown this ad.

Pretty weird that all the sports economists said that it was the pay CAP in college sports that locked in competitive imbalance but everyone else said unlocking it would make things worse. And yet we're listening to those same "everyone else"s tell us college sports is dying.

So after getting bad-AI shoved down my throat in all sorts of places, I have stumbled onto some AI uses related to board-games-online that are actually not shit. And so I am pensive about things. I will share in this thread, slowly b/c I am doing 2 other things at once.

View full thread

To be clear, I think we are on a horrible trajectory now, where AI is being used where it still sucks and also that if AI gets better that our government and society are so fucked uo that we can't possible navigate a pro-humanity solution. We are on the nightmare path in many ways.

But these new AI uses have me more worried that we might actually need to try to get off that trajectiry because the potential for mass replacement of human labor with "compute" seems closer this month than it did last month. I have no answers. But now I think we need them more. /fin

So after getting bad-AI shoved down my throat in all sorts of places, I have stumbled onto some AI uses related to board-games-online that are actually not shit. And so I am pensive about things. I will share in this thread, slowly b/c I am doing 2 other things at once.

View full thread

Maybe because I can think of exactly what I would do in my life if/when I am able to retire and focus on my hobbyist interests, it strikes me as being something a well-functioning society [ha! as if!] could navigate towards, i.e., a new basis for rationing scarce goods that took into account that...

...human labor wasn't always needed/usful, and so labor was a bad rationing tool. But several of my classmates foresaw a world where most people lacked purpose if they could not work. Am I being starry-eyed or are they just not able to imagine the sort of transformation that COULD happen?

So after getting bad-AI shoved down my throat in all sorts of places, I have stumbled onto some AI uses related to board-games-online that are actually not shit. And so I am pensive about things. I will share in this thread, slowly b/c I am doing 2 other things at once.

View full thread

I tried to at least suggest it **could** allow us all to pursue passions without worrying if we were any good at them or if they were remunerative. Obviously, we could also all be enslaved like those pod people in the Matrix and just given bread and circuses to keep us docile.

Heck, our tech overlords could decide to kill us all off save for a few millions to keep the machines running. There are lots of bad outcomes. But my classmates were focused on the boredom and the lack of purpose, not the potential for enslavement.

So after getting bad-AI shoved down my throat in all sorts of places, I have stumbled onto some AI uses related to board-games-online that are actually not shit. And so I am pensive about things. I will share in this thread, slowly b/c I am doing 2 other things at once.

View full thread

Anyway this was all in my mind last night when my Italian teacher asked us to talk about a topic on our minds. I tried to think of something less fraught, but could not, so in my horrid Italian I tried to talk through the issues in this thread. When I laid out the socialism implicit in my thesis...

... and explained I thought this could be a dream (un sogno) or a nightmare (un incubo), I got a lot of pushback from my classmates, who only saw the nightmare. To wit: If everyone has basic income and their work isn't NEEDED for society to thrive, what will they do all day? Vege out? Be bored?

So after getting bad-AI shoved down my throat in all sorts of places, I have stumbled onto some AI uses related to board-games-online that are actually not shit. And so I am pensive about things. I will share in this thread, slowly b/c I am doing 2 other things at once.

View full thread

Until now, all of the AI work I saw was so bad I was like "um, no, this is not replacing what I can do." And in my day-job, it is still VERY NOT READY to replace me. But if it is ready to replace a proofreader or a game coder, maybe in 2 years it could replace me. if not me, then others for sure.

So this is what I am driving at. How to we restructure society to absorb all of this human talent if the scope of useful/cost-effective use of human talent shrinks dramatically. I think we need to replace how society rations out income if human talent/labor is not generally economically desired.

So after getting bad-AI shoved down my throat in all sorts of places, I have stumbled onto some AI uses related to board-games-online that are actually not shit. And so I am pensive about things. I will share in this thread, slowly b/c I am doing 2 other things at once.

View full thread

would be slower or more expensive, but not so slow and not so expensive that in a world w/o AI, the project would not happen. As AI gets better (and less resource wasting?), more and more of these tradeoffs may arise, where people get replaced, not b/c of IP theft but because the machine *works*.

This latter issue, where a machine can do the work a human used to do, better/faster/cheaper is an old pre-AI problem. But it feels like there is a new aspect to the AI version, which is that all of the jobs humans developed to replace grunt agriculture and manual labor are riper for replacement.

So after getting bad-AI shoved down my throat in all sorts of places, I have stumbled onto some AI uses related to board-games-online that are actually not shit. And so I am pensive about things. I will share in this thread, slowly b/c I am doing 2 other things at once.

View full thread

First off, with the game implementation, I can't tell if the code is horrible, etc., because I only see the output, which to my non-techical eye, looks fantastic. I also can't see the energy costs, environmental costs, etc. But what I can see, as an economist, is that this could replace a person.

In some cases, perhaps, the costs of using a person to do the work would never justify the benefit, and so maybe it's AI (and all its costs) or nothing, and thus it's additive as long as the costs prove worth it. In other cases, someone would be willing to do the work cost-effectively. Perhaps it...

So after getting bad-AI shoved down my throat in all sorts of places, I have stumbled onto some AI uses related to board-games-online that are actually not shit. And so I am pensive about things. I will share in this thread, slowly b/c I am doing 2 other things at once.

View full thread

In neither case was any artwork being generated by the Ai system (in one, it was just an analysis of my own IP in a PDF, in the other it was using premade art with permission of the designer) but obviously whatever it was trained on was originally human made. Int he case of the logic checker...

I did not feel I was engaged in theft. It felt like a spellchecker on steroids. In the second case, I can easily see how it could be a tool for theft, though the specific way it was used (with permissions, as a demo to get more permissions) struck me as tolerably ethical. But here's the thing...

So after getting bad-AI shoved down my throat in all sorts of places, I have stumbled onto some AI uses related to board-games-online that are actually not shit. And so I am pensive about things. I will share in this thread, slowly b/c I am doing 2 other things at once.

View full thread

The second was that I worked with someone who used Claude Code to help a game designer (with a puboished game) make an alpha version of his own game, to show his publisher as a pitch for using BGA to revive stalled sales. I was just a play tester in the private alpha. I was very impressed by it.

In everything I've encountered vis-a-vis my day-job writing economic analyses in the context of litigation, the AI output has been utter dreck. But in these two instances, I really feel the AI did as good or better job than a human, either in terms of persnickitiness or speed of implementation.

So after getting bad-AI shoved down my throat in all sorts of places, I have stumbled onto some AI uses related to board-games-online that are actually not shit. And so I am pensive about things. I will share in this thread, slowly b/c I am doing 2 other things at once.

the first was essentially a rulebook debugger. I threw the rulebook to Skybirds into an AI system customized to assess the internal consistency of rules, and it did a good job of finding things like inconsistent use of (and capitalization of) keywords, omitted (or implicit) corner cases, etc.

It wasn't perfect, but it was better than the folks who'd done the same task for me manually. And I hd gotten to the point where I could not easily debug my own rules having read and re-read them too often.