I suspect dudes are particularly obsessed with building AI as part of their procreation drive, which is probably accelerated by women not wanting to have children (modern men all else equal would make the same choice tbf). If we can't make people, we'll make people-like things.

We can still run an LLM scorer but I like the simplicity of keeping the tooling something we can uniquely do. It also lines up with things we need to do anyway.

I’m always looking for these “double entendre” solutions - meaning one solution that solves a few problems (even partially) for us.

That would be cool. Nontrivial to make it real time, but cool.

The temptation to go broad rather than deep shows up in feature design too it seems. Go deep lads.

I thought an LLM might allow us to reintroduce freebies in a better way, but once again, I think sats are the right answer: #1421787

That was fun to explore but I think it's overkill. Plus, now that I think about it more, I feel like we can do most of what this would with satssssss:

1. a better set of stacker configured sats filters
   • a configurable filter for each posts and comments
   • can be negative if a stacker wants (effectively putting them in wild west mode)
2. territory configurable sat filters for hot/top
   • the posting/comment costs already control appearing in recent (and at all)
   • i've noticed low zap content showing up in hot of infrequented territories
3. remove the concept of outlaws and one click outlaws (currently called moderation in territory settings) from territories

Visibility continues to be dictated by whether an item has reached an investment threshold and is configurable. This just makes the floor of default visibility configurable by stackers/territories.

It's easier to explain/understand/implement.

and there's no reason to use sats

If the receiver can receive sats (horse by name), you'll send sats (from alby in your case) so that they receive sats. The portion of a zap that goes to the rewards pool (30%) will come from CCs if you still have CCs left.

Zaps are the one exception to "we use CCs first automatically," because we all want to zap and be zapped sats.

i win! i have no pre-college education and it shows.

I like the net-downzapped idea.

I did a small revamp of the wallets flows, then spent most of yesterday troubleshooting Alby's lightning address rate limiting us (seemingly due to a generic user-agent header) and WoS's implementation of Spark (or Spark itself) producing un-payable invoices from LND v0.20.0 (which the LND team fixed by contacting Spark).

Today I'll be QAing the new editor stage, reviewing/QAing a small refactor that I vibed, and continuing work on "onboiling" (which is what I call the onboarding related work).

No profile outlawing. We used to have a mechanism like that though.

I'll cook something up.

Other allegedly damning documents showed Meta seemingly bragging that “teens can’t switch off from Instagram even if they want to” and an employee declaring, “oh my gosh yall IG is a drug,” likening all social media platforms to “pushers.”

Similarly, a 2020 Google document detailed the company’s plan to keep kids engaged “for life,” despite internal research showing young YouTube users were more likely to “disproportionately” suffer from “habitual heavy use, late night use, and unintentional use” deteriorating their “digital well-being.”

All of this feels like old news, but it's a painful reminder of how humans can normalize anything if their tribe is big and they are isolated from consequences.

No one should take your word for it.

Your X account looks like an absolute scam. If it isn't, I'm sorry for saying so, but I've seen nothing from you that suggests you're legit.

https://x.com/ElevateMining

Be cautious. It's very popular to scam folks into paying for free miner shipping that they never ship.

I found someone GIVING away some old miners for the cost of shipping.

It took me over a week to determine that this was a scam, despite the culprit continuing to try to sell me expensive miners for much larger sums of money.

It does sound like a model orienting itself for a reply.

(I tend to draw pretty heavily on this finding from image diffusion models to understand how LLMs build coherent output. I'm probably overgeneralizing it.)

True, but I'm hoping to that avoid that kind of arms race by using one of these black boxes. Bayesian filters would probably do most of the work I need and much more cheaply though.

You read so model like sometimes it trips me out. Someday we'll be able to look into the models and see all the SimpleStacker weights.

The ambiguity in comparing those model outputs highlights an important point in this discussion

"llm_echo_probability": 100

brehehehheh why

I only watched the first five seconds, but my answer is that I'd have you clean my place. It's probably only an hour of work, so the other 23 hours you can watch me stare at a laptop.

hotdog or not hotdog, assmilking or not assmilking, is approximately good enough for what I'll need initially so I could start small.

afaict most of the trouble with this stuff is the non-model parts. still, this thread has already proven useful and the thread is young as they say.