I have also done similar research because I wanted to build something to handle microtransactions on a personal website that could scale if adopted to be usable by everyone if they wanted.
I looked at crypto currency because it seems like the obvious naive solution. it doesnt work. the cost of the transaction itself far outweighs the value of the transaction when dealing with fractions of a cent. you want an entire network to be updating ledgers with ~millions of records per ~$1000 moved. the fundamental tech of crypto leans towards slower, higher value transactions than high volume, small transactions. Lots of efforts have been made with some coins to bring down the bar of "high value, low volume" to meet everyday consumer usage rates and values - but a transaction history at the scale of every ad impression for every person is a tough ask and would perpetually be in an uphill battle against energy costs.
Ultimately, the conclusion I came to is that the service would need to be centralized, and likely treated as cash by not keeping track of history. Centralized company creates "web credits", user spends $5 for 10,000 credits, these credits are consumed when they visit websites. Websites collect a few credits from each user, and cash out with the centralized company. The issue is that since it would cost more to track and store all the transactions than the value of the transactions themselves, you have to fully trust the company to properly manage the balances.
I started building it and since I would be handling, exchanging, and storing real currency - it seemed subject to a lot of regulations. It is like a combination bank and casino.
i've thought about finishing the project and using disclaimers that buying credits legally owes the user nothing, and collecting credits legally owes the websites nothing, and operating on a trust system - but any smart person would see the potential for a rug pull on that and i figured there would not be much interest.
The alternative route of adhering to all the banking regulations to get the proper insurances needed to make the commitments necessary to users and websites to guarantee exchange between credits and $ seemed like too much for 1 person to take on as a side project for free
It would need to be mostly centralized, but keeping track of history would not be hard.
A typical credit is getting paid in, transacted once, and cashed out. And a transaction with a user ID, destination ID, and timestamp only needs 16 bytes to store. So if you want to track every hundredth of a penny individually, then processing a million dollars generates 0.16 terabytes of data. You want to keep that around for five years? Okay, that's around $100 in cost. If you're taking a 1% fee then the storage cost is 1% of your fee.
If your credits are worth 1/20th of a penny, and you store history for 18 months, then that drops the amount of data 17x.
(And any criticisms of these numbers based on database overhead get countered by the fact that you would not store a 10 credit transaction as 10 separate database entries.)
fair enough on tracking history in the centralized model. I had suspicions there would be hidden costs that might make it too expensive. i dont think the data storage would be as much of a problem as the cost to write it to storage.
I wasn't fully envisioning credits only being transacted once before cashout either. I was thinking more along the lines of being able to create something that goes viral, a lot of people use it and you rack up a bunch of credits, and then you can sit on those credits and spend them as you use the internet without ever having to connect to a bank yourself. So people who are contributing more than they are consuming would rack up credits. they could use those credits to enrich their contributions, maybe pay for cloud services, etc.
the credits could form its own mini web economy if it got popular enough. As cool as this would all be if done honestly, I know that if i saw a company telling me to buy web credits to use anywhere on the internet and the websites get to decide how much to charge and they charge it automatically when i visit the website, and if the company i buy the credits from goes out of business then i may not be able to cash out or get my money back, then I likely wouldnt be buying those credits... so idk
Even with user to user credits it would take a lot for the number of transactions to go above 2. That would mean more than half the money is going to viral payouts.
And was this assuming you'd only take a cut on the cash going in and out? Because even a 0.1% cut of the transactions would mean you have $1000 to handle the amount of data I described in the last comment.
>And was this assuming you'd only take a cut on the cash going in and out
I think fee needs to be per transaction, maybe not cash flowed per transaction but accrued per transaction.
Say we both self-host a website for our favorite daily game, and I use yours about as much as you use mine. We would transfer roughly the same amount of credits back and forth to each other ad-infinitum. but the credit service provider is accumulating only expenses with each transaction.
Say someone make a lot of bot accounts to simulate user traffic, and it sends each of them credits to use to visit their own site. the host collects the credits from the bots and transfers them back to the bots to keep them running.
Late reply, but in case you'll check I think most of what I said is sourced from sources of varying quality and salience, but at least it's sourced from somewhere. But I just typed it all out quickly without checking anything over, so a lot might be wrong. But it's not entirely pulled out of my ass at least.
Evolutionary history is of course always difficult. I think the loneliness part comes mostly from the kurzgesagt video on loneliness, as well as some other stuff here and there. Rate of infanticide is roughly correct with quick Google. Rest of tribal stuff is from a variety of books and high school social anthropology. I think I actually have the "reasoning for infanticide" part from sex at dawn, of all places.
I'm always scared to run a deep research service to find the counterpoints after I type this kind of stuff out, but feel free to do so for me and dress me down. At least survivorship bias is a classic that's pretty much always worth keeping in mind on any topic.
without corruption you could do a shitty job once and then you won’t get another contract because you did a shit job
with corruption the quality of the work won’t matter so in the extreme case you can deliver nothing at all and you’ll still keep getting contracts - In my country we call this being “plugged in”
i love this and have been a long time complainer that browsers dont automatically operate this way.
how does it handle forms or homepages with refreshed content? for example, the home page of hackernews - will it always show the latest feed from the last time i had a connection or will it store each time ive visited it ?
I do not trust any LLM. But I am with the other person, the intention is not to discredit you or make you convince us - you are doing exactly what a comment section is designed for. Your comment is so good, in fact, that we want to trust it more than a comment in general deserves to be trusted.
while i agree its better to go off and prove it to ourselves, there is merit in having a conversation here
how is it a conflict of interest for a google product to have a bias towards using google products?
As users we must hold some accountability. AI is aiming to substitute for humans in the workforce, and humans would get fired for recommending competitor products for use-cases their own company is targeting.
If we want a tool that is focused on the best interest of the public users, then it needs to be owned by the public.
"Conflict of interest" isn't exactly the right term. "Conflict of value proposition" perhaps? E.g., you're using Google search based on the proposition it will effectively find things for you, but that turns out to be not what it actually does.
First, I've spent a ton of time becoming opinionated about a normalized data model that supports the product experience I'm trying to build. This applies both to the extraction (line items, warranty sections, vendors, etc.) and the analysis portion. The latter is imperfect, but aligns philosophically with what I'm willing to stand behind. For example
- building outputs for price fairness (based on publicly available labor data)
- scope match (is vendor over/under scoping user's intent)
i work in cost and pricing, and while i see the allure of AI helping out with it and I would love to be able to hand it over and work on other things, i feel like so much of this work involves things outside the sandbox.
Take price fairness, for example. i feel like the human part is core to this work. it ultimately all comes down to a test of reasonableness. A wide brush for costs is sometimes used because it keeps things trackable by the humans involved. An AI is able to generate an amount of work thats unreasonable to verify. At the end of the day, pricing is a negotiation not a logic puzzle.
If it does work though, i think it could open a huge door for Cost Plus Fixed Fee contracts which seem like the fairest contract type but often come with too much burden of paper work compared to the more popular firm fixed price option
I agree. Additionally, a company can own and update a business language of their own design at their own pace and need. Then they can use AI to translate from their controlled business language to the DSL needed (translation being an area it actually does well). In this way the LLM would only ever be going from General -> specific, which should keep it on the rails, and the business can keep its business logic stored
Now that said, there is still the actual engineering problem of leveraging the capabilities of the underlying technology. For example, being able to map your 4 core program to a 16 core system and have it work is one thing, actually utilizing 16 cores is another. Extend to all technological advancements
> I agree. Additionally, a company can own and update a business language of their own design at their own pace and need.
Yes, although I was more thinking of this being in most cases a SaaS offering because the implementation of the DSL needs solid non-LLM engineering. Larger companies will be able to afford an internal platform team, but most won't.
> Now that said, there is still the actual engineering problem of leveraging the capabilities of the underlying technology. For example, being able to map your 4 core program to a 16 core system and have it work is one thing, actually utilizing 16 cores is another.
I see this more of an extension of existing trends, for example Wordpress themes with limited customizability. Most DSLs won't allow full utilization of the underlying technology, on purpose because that's the only way to keep it simple. I do see this leading to a split into two classes of developers: those who only target simple DSLs using an LLM, and the "hard" engineers who might use LLMs every now and then, but mostly not.
I see the angle you're coming from now, more mass market and expanding best practices from bigger companies out to medium and small businesses looking for plug and play solutions.
I was thinking more about what I believe you describe as the "hard" engineers, and would say the power AI provides for mapping and translating will greatly benefit those teams as well with the right set-up. People are pushing for the "code for me" angle, but i think there will be a lot of opportunity to have LLMs take on a middle ground of syntax management while the engineers manage the system effects. for example, the engineer may be deciding whether to use a linked list or binary tree and the LLM is implementing it with the available code stack approved by the company.
A company that can successfully implement such an LLM opens up their talent pool from people who know their stack (or want to learn it) to people who know any stack
> for example, the engineer may be deciding whether to use a linked list or binary tree and the LLM is implementing it with the available code stack approved by the company
At this point it's a slightly more sophisticated version of the IDE's "refactor tool". If, in addition to replacing "HashMap" with "LinkedList" in a bunch of places, it might also fix tests, then it's indeed useful but won't be worth paying much more for it.
> A company that can successfully implement such an LLM opens up their talent pool from people who know their stack (or want to learn it) to people who know any stack
Think about it: if the business usefulness of a tool is mostly in reducing onboarding time by even a 75%, it's not really that valuable.
I looked at crypto currency because it seems like the obvious naive solution. it doesnt work. the cost of the transaction itself far outweighs the value of the transaction when dealing with fractions of a cent. you want an entire network to be updating ledgers with ~millions of records per ~$1000 moved. the fundamental tech of crypto leans towards slower, higher value transactions than high volume, small transactions. Lots of efforts have been made with some coins to bring down the bar of "high value, low volume" to meet everyday consumer usage rates and values - but a transaction history at the scale of every ad impression for every person is a tough ask and would perpetually be in an uphill battle against energy costs.
Ultimately, the conclusion I came to is that the service would need to be centralized, and likely treated as cash by not keeping track of history. Centralized company creates "web credits", user spends $5 for 10,000 credits, these credits are consumed when they visit websites. Websites collect a few credits from each user, and cash out with the centralized company. The issue is that since it would cost more to track and store all the transactions than the value of the transactions themselves, you have to fully trust the company to properly manage the balances.
I started building it and since I would be handling, exchanging, and storing real currency - it seemed subject to a lot of regulations. It is like a combination bank and casino.
i've thought about finishing the project and using disclaimers that buying credits legally owes the user nothing, and collecting credits legally owes the websites nothing, and operating on a trust system - but any smart person would see the potential for a rug pull on that and i figured there would not be much interest.
The alternative route of adhering to all the banking regulations to get the proper insurances needed to make the commitments necessary to users and websites to guarantee exchange between credits and $ seemed like too much for 1 person to take on as a side project for free
reply