pull down to refresh
21 sats \ 0 replies \ @fourrules 26 Nov \ on: Did the stacker news PWA forsee Google’s tightening noose? devs
Redeveloping a web app for a native environment should become trivial with LLMs writing most of the code. I feel like maintaining multiple client apps doing effectively the same thing will be relatively more straight forward and the economics will make more sense.
So you're basically describing a circular economy. That's not radical, it's just very difficult and still a collective action problem.
What you could do is create a Chrome extension that converts all prices into bitcoin.
But that will only impact the tiny percentage of Bitcoiners who hear about it, care to install it, and then overcome the inconvenience to continue using it. And then what? It's not the sellers.
Convincing non-Bitcoiners to price their goods and services in a currency they don't accept? This seems like a waste of time. What am I missing?
I think this is the last phase because it's a collective action problem. Bitcoin itself didn't have that problem because it advantaged early adopters.
How do we advantage early adoption of bitcoin as unit of account? It's not going to be "why don't we buy more stuff from businesses that use bitcoin as unit of account", because that just moves the collective action problem, it doesn't solve it.
It especially cannot happen if so much is short term demand comes from leverage, or if the bitcoin mining industry is leveraging up to convert to AI data centers, which creates volatility.
This is a conclusion I came to in 2017 and have been trying to solve it since.
Unfortunately I don't have the expertise to implement a protocol that would decentralise reputation. What Bitcoiners don't want to talk about it is reputation and trust are two sides of the same coin, and storing reputation data anywhere is fraught with risk and potential abuse.
My instinct is that trust has to start from hyperlocal networks below the Dunbar number, consider them tribes, which themselves have reputation and membership confers on a person the status of the tribe onto the member. Maybe you could have even smaller networks inside tribes called "families" or "projects", and then above twice tribes "communities" and "societies".
Reputation can be governed by an algorithm modelled on Pagerank with different weighted endorsements, on a scale from implicit to explicit, or increasingly costly, e.g. follow = implicit endorsement; donation = explicit endorsement.
Maybe something like this can be built onto Nostr, where there is a concept of continuity of identity that is unlike bitcoin.
Either way it needs to be built out of a system that aggregates real world event data.
Someone working in the ad business pretending to have a conscience because he doesn't like changing his workflow.
These examples aren't comparable. They are all used by individuals within organisations, not by organisations coordinating with other organisations.
And despite that fax lasted for years, I'm sure it's still used.
For bitcoin to be used for corporate settlement you need to coordinate with and get sign off internally and externally, and there are always intransigent people, often at the top, who will block it.
Of course it's probable that Russian corporations today are using bitcoin for corporate settlement, since the sanctions would have provoked an environment for experimentations, but they've probably migrated to the Chinese payment rails since then.
I'm sure they can make a decent estimate from total nodes and downloads of their OS, along with their own Start9 sales and the number of support requests they get for each OS, along with some other reasonably proxies.
They would track that as a KPI internally, and therefore would be incentivised to be reasonably objective, even it is ultimately an educated guess.
Whether they published the same number as they have estimated internally is another matter.
But no spying is required for an analysis like that.
The underlying issue is the idea of revealed preferences, that underpin economic theory and justify all of these systems. "Give the people what they want" is what the tech bros say to sooth their conscience so night.
This is spot on. The closest we will ever get to UBI is COVID stimulus, and we saw the inflation of asset prices and consumer goods. If AI takes most jobs it is more likely to end our system of government than to result in any kind of basic income.
Democracy is a temporary interlude, ye interregnum where civil war is rendered non-violent, licensing a group to enact violence on behalf of the state on citizens in order to maintain the illusion of peace. In a majoritarian democracy one vote is a proxy for a gun that would be used if the democratic peace were to fall. Women are given the vote because the men with the guns need them and they therefore have influence. The old and infirm have votes because they have experience, influence, and other skills in case of renewed violence.
But now we have consensus collapse so the peace is unstable. If jobs are decimated then the peace will unravel, as states will no longer have a monopoly on either the monetary system or violence, rendering them entirely redundant.
In the case of a civil war you need to pay fighters, but you can't pay them in a fiat token if they are not certain that you will win. You need to pay them in something that they can still use in case you lose and they have to escape. So if a UBI is distributed it will just drive up the price of an equivalent digital token that they have complete sovereignty over.
In the case of violence all of the weapons come from other countries where they are manufactured. Literally anyone can access cheap weapons now, whether from Russia or China or Iran, given a sufficient degree of organisation, which is granted by the stability of the non-state monetary system.
So if the "far right" or any unsanctioned parties were to legitimately win but not allowed to govern they would have absolutely no reason not to turn over the table and declare the end of the democratic experiment.
The best case scenario is a distributed participation income system that makes it worthwhile for the people with the capital to put down their guns and accept a new model for the welfare system that is market-based and not administered by a self-interested bureaucracy.
The bigger problem is that it would take a full year for everyone to migrate to the new quantum resistant wallets at current block size limit, so even if there is just-in-time forced consensus on a solution, it's still going to be too late to avoid chaos.
We need to be more or less agreed on a uncontentious and well tested solution 10 years before a quantum breakthrough.
Thankfully there are lots of eye balls in this that are not those of Bitcoiners, so as long as most big breakthroughs are published and widely propagated it'll be well telegraphed.
Still, plenty of scope for everyone to accuse everyone else of being a psy-op agent.
I agree with that over the 50-year horizon, but LLMs aren't yet having the degree of impact that white collar workflows. You talk to your average Salesforce employee and they're barely using LLMs at all. The job cuts and hiring freezes are being done:
-
because interest rates make companies motivated to reduce headcount to head off anticipated tighter margins in a stagflation environment
-
in anticipation of potential automation opportunities and increases in productivity that may allow them to increase output and throughout with a lower headcount
-
because LLMs are a convenient means of encouraging current employees to watch their backs and up skill rather than complaining about increases in workload due to reduced headcount. In other words "if you're demotivated by the layoffs you're just not hyped enough about the productivity gains provided by LLMs, and you're not a good culture for anymore"
I use LLMs every day, but they're not yet penetrating the workflows of an average white collar worker enough to justify the conclusion that they have such an outsized impact on job openings on their own.
Its absolutely the latter. Companies have been trying to get lean without demotivating the staff that remain. They strategically lay off a bunch of people just to see who was making a difference, who the remaining staff fight to get back, then they re-hire discriminately. They were doing it way before OpenAI entered the popular consciousness.
But that doesn't preclude the long-term trend that LLMs have accelerated, that automation is replacing jobs. If interest rates were lower it would mean companies would be hiring to create new projects that leverage AI, increasing head count and bloat, the narrative would be totally different, shaped around opportunities and new paradigms. But because there is motivation to shave some pounds off the AI boom is manifesting in brutal job cuts and hiring freezes.
I think the strategy is not dissimilar to that of Berkshire Hathaway, difficult as that may be to believe.
Step 1: accumulate bitcoin
Step 2: wait for credit crunch
Step 3: offer collateral for bridge loans in return for a cut of the interest and potentially equity in the distressed companies
In the case of Berkshire they are waiting for the same blood in the water, the same credit crunch, to use their cash to buy up distressed but viable assets.
During a credit crunch, ironically, the value of bitcoin as collateral increases because what you want in collateral is upside potential. Given the quantity of bitcoin that Strategy holds even in a downturn he'll be able to offer collateral to a lot of good businesses in distress, especially technology companies because of the industry expertise he has in-house, something banks often struggle with.
In that context one would expect bitcoin treasury companies to pop up throughout the economy, randomly distributed, with local industry relationships that they'll be able to leverage in the same way, to form partnerships with banks who want to provide credit but need collateral to back it. That's what we see.
I have some ideas:
-
pay-to-post: anyone can add a business, but you drop sats to do so, and if the business gets a lot of engagement from other users they're likely to zap you, so you get your money back. Confirm the business details with Foursquare API, same data as OSM.
-
Users can comment in case a listing isn't valid anymore, and there should be a simple traffic light system that indicates whether they still accept bitcoin
-
Add a gift voucher feature with an escrow service. Users can post business who don't accept bitcoin but who do sell gift cards online. User-A wants to spend bitcoin in that place, they drop an anonymous request for someone else to buy a gift voucher for a specific value, giving an email that they should receive the voucher at (can be burner, not displayed publicly). Users-B sees an opportunity to buy a small quantity of non-KYC BTC, so they buy the voucher and add the email they are given for the request (not the actual email of the person selling bitcoin), something like voucher+644775sgg@satmap.org. The voucher is held in escrow at that email. To receive the bitcoin User-B creates an invoice for the amount and provides the payment link (system could do this automatically), and that's what is given to User-A. When User-A pays the bitcoin that triggers the escrow service to send them the voucher email. Some LLM can run a check on the email possibly, to ensure it is genuine and contains the code.
The primary problem to solve would be ensuring people are not spoofing the voucher emails. They can only be claimed once very often, so you can't have someone in background spot-checking that the vouchers purchased by new users are valid. You need some kind of reputation system that encourages repeat interactions, much like classified ads websites and apps. That means you need a request-accept system, so User-B has to request to purchase the voucher, and when User-A accepts then User-B sees the email address to send the voucher to.
Problems:
So now User-A has agreed to the exchange with User-B, and the system will know when User-A pays out the bitcoin, but:
- the system won't know whether User-B has actually taken the bitcoin AND spent the voucher themselves before User-A has had a chance.
- the system won't know for sure that the voucher received is real
- User-A might accept, prompting User-B to purchase the voucher, only for User-A to ghost them. In this instance the voucher is still valid and held in escrow, any other user could pick it up by paying the lightning invoice, which would become visible to other users after a holding period for User-A to conclude the transaction.
I can't really identify any other problems to solve beyond basic UX. Each user can review each other once an exchange is accepted, so aside from that you just have the basic cold start problem of nobody having any reputation. Usually the cold start problem is solved by users repeatedly engaging with the site for another reason and eventually a sufficient number of users take the leap and start building their reputations. New users just have to start with very low denomination voucher requests.
Oh, another problem is density. If I want a restaurant voucher I want it now, not in 3 weeks. But in principle I think the system could work with a small number of bitcoin-buyers and a high number of sellers.
If you got that system going then the whole community would be able to identify businesses that Bitcoiners like but that don't yet accept bitcoin directly, so that we can target them through repeat custom and engaging with managers, show them how popular their business is with Bitcoiners and help them to get onboarded as bitcoin merchants.
I think this system would be incredibly popular in Italy where you have a lot of restaurants and bars, lots of tourists who want to spend bitcoin without CGT, or locals who cannot draw down without declaring and paying a huge amount of back tax. In Italy taxes on crypto are brutal.
These proposals are theatre. They seem shocking to people outside France, but the truth is that by the time they are implemented, if ever, they are so well forecasted that almost everyone who would be impacted have deployed defensive measures, which is exactly the intention of the people creating the law.