mshumi

I think the most interesting thing about making AI to serve human interests is that ultimately the system of AI (foundation model + infrastructure + energy) should resemble a human system.

Well done, this is a really cool approach to leveraging ChatGPT. Messing around with it locally now

I think this article is particularly relevant after SVB. Also the large interest rate differential on most American checking accounts vs FFR.

X, Y, and Z Are NYTimes Poisoned

Buffer of batteries and generator backup

To encode NOT something takes one more bit than encoding something.

Zero real rate of return. If bonds have a negative real rate of return and stocks have a RISK-ADJUSTED negative real rate of return, then gold or more generally commodities are a better option.

Judging by the lack of a market reaction this morning, this is mostly immaterial.

Take this line of reasoning further and you reveal the problem with crypto to crypto taxation.

BTC Ledger =/= BTC consensus. If the protocol can be made pure proof of stake while maintaining the same availability and security guarantees this is no longer an issue. However, this is against the best interest of miners.

Unfortunately, aerospace isn't an industry where you can say "fuck it, ship it"

In many of these discussions why do we always refer to government as an abstract "they", why is it never "we or us". I think we've lost a connection between services provided by government for the public good and the means to pay-in for those services. Historically you paid taxes in real value (grains or gold) in exchange for protection, enforcement of property rights, and public goods. We do the same today, but the current state of income taxation seems wildly disconnected from value provided by government. Personal opinion is that LVT, VAT, excise taxes, and taxation in equity is a better way to link equitable distribution of real value to benefits.

We could easily solve this BTC problem by hard forking to a pure proof of stake consensus protocol for BTC. Same security and decentralization guarantees. No environmental harm, and far more scalable so you can actually buy coffee with it. Ofc that's against the best interests of miners and BTC core devs.

Fork BTC to Pure Proof of Stake consensus. Balances mapped 1:1. No Environmental problems, no scalability problems, no latency problems. It's not a question of if this fork happens, but when.

One wonders if it is some fundamental assumption that limits further understanding of the Universe. Is it our intuitive relationship with numbers, time, and physical space that is limiting? We all grow up with societally imposed relationships between numbers, time, and physical space, but are these learned or do they actually reflect objective reality. The frameworks derived from these fundamental assumptions are excellent for making predictions and are testable within those frameworks.

Hard fork to pure proof of stake, problems solved. However, against Miners' best interests

underwriting cap and dischargeable in bankruptcy would do it

Have you considered the possibility of a crash up in asset prices (currency devaluation, negative interest rates). Cash wouldn't be so good in that scenario. I feel similarly about current market valuations and mostly hold cash-equivalents, however I've also bought a significant amount of otm call leaps on agriculture/metals funds to hedge a crash up scenario.

It's more likely to be a crash up than a crash down. Interpret that as you like.

TIPs adjusted by cpi determined by gov. Better off holding real assets. For retail, a large mortgage is a fantastic bet on inflation. For the more risk-loving, long gamma on any vehicles tracking real assets has really good risk/reward in a high inflation market.

BTC fork using Algorand consensus. Same bitcoin, same LEDGER, but the consensus technology is vastly superior. Why willfully ignore 10 years of advances in cryptography and distributed computing.

Hardfork BTC to use Algorand consensus. Put stablecoins on the same chain. See the magic happen.

A hardfork to algorand consensus would solve this

It should be possible to have similar UX/Support to traditional banking while having users custody their coins. Multi-sigs and digital identity would be very useful in this regard. Trusted 'recovery' parties or hosted services that intentionally limit risky transaction behavior. For example, you could have a scheme where funds in a smart contract are gated by a personal signature allowing one to spend up to $7500 in a 24 hour period, but requires secondary approval provided by a trusted company to send more (on top of your signature). If you need to break the limit there could be a verification requirement from the company or emergency override key you keep safe. Crypto will be most successful when people do not even know they are using it. Assuming no legal issues with adoption, there are more challenging taxation and liquidity considerations to using fixed supply currencies at scale than usability for the average non-technical user

Bootstrap it by writing bots with believable names. I believe a lot of .io games use bots in place of players when there are not enough players online.

It seems absurd that transfers between two accounts you own at large financial institutions take 3 days to process. There are obvious safety concerns, but there should be provably safe and faster ways of transferring value at minimal cost. People should be able to specify their security level; you could even gate-keep it with licensing, but we should make it an option for people.

I'd happily buy pizza with SPY. Why keep money in a near zero-interest bank account when I can hold beta. If markets were 24/7 with instant settlement of securities I could hold all my money in stocks and bonds. I wouldn't be concerned with holding cash for liquidity's sake. More so I could hold more of my money in securities representing interest in real property -- housing, arable land w/ fresh water, and commodities which are usually a better store of value than cash unless there is a liquidity crunch. There are good reasons why asset prices have been soaring for the past decade.

Crazy idea in this space. Eye-tracking will eventually be standard on consumer VR headsets that can be had for $300. As someone following the industry closely, I believe that the Oculus Quest 4 or a device like it will certainly have eye-tracking support. Obviously if one were to take their test in a VR environment with their eye-movements tracked it would be almost impossible to cheat. Eyes also provide a good biometric identifier. Mirrors, screens, exploits, etc can be foiled using software locks, HMDs cameras and imu data. Of course I'm not advocating a testing system like this, but it would put a hard stop to cheating on synchronous exams, enforce some level of standardization, and should be accessible to every student in the 1st world. Understandable privacy and ethics issues with something like this. It is probably better to address why people cheat in the first place and make education less reliant on exams. (continuous low-impact variable feedback vs discrete high-impact fixed feedback)