JFC, people in tech are really out there saying that language models will be better at therapy, financial advice, and career advice than trained people.
WTF is wrong with you people? Do you really have no clue about what other peopleβs jobs actually involve?
Language models canβt even do maths how are they supposed to get good at financial advice?
And therapy? Justβ¦ π
Whatβs wrong with people in tech?
@baldur I thought all these use cases were already covered by blockchain?!
π
@enoch_exe_inc by that, do you mean that distributed ledgers are a bogus idea, or that only blockchain is a bad realisation and there are better ones, and if so, could you point me in that direction please?
@AeonCypher @tivasyk @enoch_exe_inc @aguleb @baldur its also "trivial" to forge a fake chain history - getting clients to accept it is the hard pard, but git clients *also* don't blindly accept forged histories - you need some way of handling fresh clones and deciding which *new* additions are valid, but for most applications you can just pick someone to trust.
the new bit was "trustlessness," but that doesn't solve a problem most (non-cryptocurrency) applications actually have
@cr8 @tivasyk @enoch_exe_inc @aguleb @baldur
You can fork a repo, rewrite the entire history, and publish it as your own.
The entire reason you can't just mess with the history is because a maintainer decides whether to bring it into the main branch.
That's not distributed. It's impossible for Github to both be distributed and verifyable. It can be only one.
Also, maybe go read the actual papers on blockchain from the 80's and 90's.
You clearly seem to have a misunderstanding of what the technology is, and continue to confuse its application under the capitalist hegemon for what the technology actually is.
Proof of Work, and Proof of Stake are not possible under any standard version control system, and AFAIK no system of proof exists for them other than through a central authority.
@AeonCypher @tivasyk @enoch_exe_inc @aguleb @baldur https://github.com/ctfs/write-ups-2014/tree/master/stripe-ctf3/level1
there wasn't a mempool or anything but if you were really into the idea of proof-of-work version control there's no reason you couldn't take it further
@cr8 @tivasyk @enoch_exe_inc @aguleb @baldur
It's hard for me to imagine a system of federated micropayments, which we're going to need as the old internet dies, without resorting to some form of blockchain.
The alternative is a central authority.
I'd love to hear alternative ideas.
Though, I'll be honest, my complaint here is with the principle of attacking the technology and not the people wielding the technology.
A technology is just a tool.
@ShadSterling there are ones where those are less bad (though not eliminated) - I think the volatility is a worse problem - its not really that useful to accept payments in a currency you can't pay your *bills* with and may not be worth the same amount by the time you *can* exchange it.
stablecoins aren't a solution either - they *are* a central authority, and algostables (or generally any token pegged to *anything* off chain) *provably* can't guarantee they won't depeg.
@gfwellman @cr8 @tivasyk @enoch_exe_inc @aguleb @baldur
How and why? - Because everything is going to be paywalled. Because AI can read things for you.
"it won't support micropayments? " - Micropayments currently exist from authorities. I said federated micropayments.
"We must keep our system of laws and government democratically accountable so that central authorities remain trustworthy"
I would _MUCH_ rather that banking, as in money creation, and for that mater payment processing were run by a competent government. I'm not holding my breath for that.
@AeonCypher @gfwellman @tivasyk I think money creation is not really a problem with a technical solution.
Decentralizing things at a protocol level *does nothing* to decentralize **power** - pretty much every chain of any size has far more drastic wealth concentration than most govt currencies and could be (and sometimes are) crashed by the actions of just a few individuals.
@cr8 @AeonCypher @tivasyk This makes the argument about power much better than I did. If we lose our democracy to a red-ballcap-wearing mob, the leaders of that mob are going to steal everything they can, like Putin-affiliated oligarchs. Federated micropayments will likely reproduce the wealth concentration problem that @cr8 described, and would probably just be taken over by oligarchs anyway.
I really do sympathize with the goal of "how do we keep our ability to conduct commerce away from central authority if the central authority goes full-fash". I just doubt that a blockchain could be kept free under those conditions.
@AeonCypher @tivasyk @enoch_exe_inc @aguleb @baldur i worked on a lot of similar pre-"blockchain" tech and then after it took off watched a lot of my ex-coworkers and peers go on to blockchain co's that I thought were at best producing clunky subpar products and sometimes just outright scams
the incentives are toxic - crypto is a space where investors don't have to wait for an "exit," and can profitably just sell hype so even well-meaning projects get turned into pump-and-dump schemes
@cr8 @tivasyk @enoch_exe_inc @aguleb @baldur
Oof, yeah. That sounds rough. I definitely get the ire, hatred even, for the crytpto grifters.
Was your project using a trust network? Most people's eyes just glaze over at the concept. LOL
(Sorry if I was a bit aggro earlier. I'm having that kind of day.)
@AeonCypher @tivasyk @enoch_exe_inc @aguleb @baldur was mostly on things closer to a "private blockchain" than public one w/ adversarial consensus (didn't have to work with *totally* mutually distrusting parties - it was still verifiable, you couldn't undetectably change history - but we never needed a solution for a "cold start")
I generally like the idea of trust networks but I think the hard part of them is UX - its hard to make clear (especially to nontechnical users) what's happening.
Baldur Bjarnason
Beluga
enoch_exe_inc
ΡΠ²Π°ΡΡ ΡΠ°ΡΠ°ΡΠΈΠΊ
β΄β΄β΄ΓπΖ.Ζκβ΄π₯ΡΙΌπ©π»βπ»
inhuman resource
ShadSterling
Greg Wellman
Stephan Eggermont