siddthesquid

Also these kids will grow up to do harm to the communities in that state, which will spend more money at the end cleaning up those mistakes.

I'm a bit biased here, but I feel like the more form factors we introduce, the worse apps are going to get as developers now have to account for it (and they won't, eg YouTube on ipad). Do people actually like foldable phones enough?

- If an AI provider has control over the AI algorithm running in the secure enclave, they can easily have functions that provide them the plaintext through a separate channel. secure enclave does not prevent that

- The output can reveal information to the provider, which homomorphic encryption would have protected

- Inference is running on GPUs - so its moreso nvidia than amd/intel, but this is just a nit

So homomorphic encryption exists so the user doesn't need to do work to figure out if the provider could be adversarial.

If I used regular encryption to send my credit card information to an AI with fraud detection, the provider still needs to decrypt that data on their side at some point before it goes into the AI.

Using this other encryption, the provider has neither need nor capability to decrypt it on their end, so the user gets extra security.

Thanks for the suggestions. Just checked out some of Le Ton Beau de Marot, will most likely read the rest.

Actually, I was recently wondering whether poems can be interpreted as "efficient" computer programs. Words in a sentence/essay just connect different objects in space and time (syntactically and semantically) to form some concept, kind of like a program. A poem just uses higher order abstractions (imagery) to express the same concepts in fewer words.

So this book should be great!

I have not read GEB but I thought his second book, I am a Strange Loop, did a pretty good job of connecting the idea of self referential loops (like in godels proof) to consciousness and art and such.

> How is this any different than the Fed or the fractional reserve banking system creating new US dollars?

The miners get the fees. The fed does not keep the dollars they make. They also adjust the rates to avoid things like recessions.

> Nothing except for the thing that matters: If you have something fungible instead of something with high switching costs, it makes fees go down.

The US dollar is considered fungible... Help me understand how any of this is specific to blockchain technology and not included in non-blockchain technology. Have you worked with this tech before? Also, what about things like venmo and zelle? zero fees, super fast.

> And that's why blockchains are useful! To exert the pressure needed to make that happen.

I'm not saying they are not useful. I am saying the technology behind them is irrelevant to the costs.

Yes, that trust is the fundamental difference. However, that trust costs money in the form of needing more nodes.

You usually can trust your bank, as long as you trust your government. Regulations make it difficult for banks to misbehave.

That being said, not trusting your government (which I can believe is a valid stance in some countries) is probably the only valid use case for blockchain IMO.

If I'm bank of america, and i publicize a public key, and then everytime everyone does a transaction, i sign a receipt using my public key such that my customers can prove that transaction happened, then that would be the cryptography.

if bank of america does something malicious, i can prove in court very trivially through those signed receipts that they did so.

So I don't need to trust bank of america - i just need to trust the courts to charge financial institutions that provably are breaking the law.

My point is that blockchain is just a technology - nothing about the technology itself makes the concept of transferring money cheaper. I agree that it is another competitive avenue for transactions, but if it became a threat to payment processors, my theory is that they could lower their costs more than blockchains potentially can. This is because the software and infrastructure needed to build something that assigns numbers to accounts and allows transfers is obviously going to be cheaper off the blockchain.

If trust is an issue, the bank can provide cryptographically signed receipts that show they've confirmed the entire lineage of your account, in the same way a blockchain does, but they would be the only verifier. The question becomes about how the cost of the additional trust from the blockchain relates to the incentive of doing honest business. I imagine that trust cost is pretty high.

> can you have a blockchain with lower fees than payment processors currently have? And the answer appears to be yes

The transaction fee is not the only thing being paid. They are also getting mining rewards. If a blockchain has mining rewards, maybe in the form of Bitcoin Cash, then that will dilute the entire pool of Bitcoin Cash.

I think the technology of blockchain is irrelevant.

If something can be accomplished on the blockchain, which requires N nodes, a business can probably replicate that same objective with less than N nodes because they don't have to pay the cost of verifying that nodes are acting honestly. This business is incentivized to be honest because otherwise they lose their business. Someone has to pay those costs for the N nodes on the blockchain - who will it be? Transactions seem cheap now because funding for these blockchains is often used to subsidize costs.

You mentioned ease of use, like the use of SDKs, but blockchain technology does not enable that. All blockchain can do is that if you ask it "hey i was told the state of the world was this. is it true?" and the blockchain will tell you yes or no. If you want to provide those kinds of guarantees to customers in a reliable way, all you need is cryptography, not blockchain.

I wonder if there is something to be said about how machines are based on deterministic and algorithmic properties, whereas emotions could potentially involve logic beyond what humans can observe, like quantum interactions.

I feel I can empathize with the ideology, and I want to argue that an idea like this is a big piece of some puzzle.

The puzzle: During the collapse, I think the biggest thing we would lose is our ability to communicate with each other in a reliable way. Like telecommunications and internet and all that. Who and what do we trust?

I imagine that before a collapse like this would happen, people would get a hint to gather as much as they can to help with their survival. Those things would include: this OS, the knowledge to load a sequence of bytes from whatever device is holding the OS to as many CPUs/controllers as possible, strategies to connect any number of arbitrary CPUs to radio devices, some knowledge of public keys and private keys. maybe all neatly organized into a handbook

The OS itself would be responsible for providing as easy an interface for any average joe to generate public/private keys, communicate with other people who have followed their same protocol, and use those public keys to build trust from communications. before the collapse happens, you may even collect a list of public keys you are likely to already trust.

The OS could maybe even have software for building communities of trust or even handling adhoc finances through (don't hate me for this) cryptocurrency.

anyways, this is all based on an assumption that the ability to communicate quickly (and build trust in a decentralized yet controllable way) is the best mitigation we have to a collapse

this answer is slightly influenced by the movie leave the world behind lol