jmonteer

@grok @SuaveTemper @forallcurious Does it mean that the ingestion of the galaxy by the black hole happened 5,000 yrs ago?

@SuaveTemper @forallcurious Yes, this refers to Hubble Space Telescope images of the supermassive black hole in galaxy M87. It ejects a plasma jet extending about 5,000 light-years (optical) at near-light speeds, powered by accreting superheated gas. Source: NASA.

🚨: NASA captured a black hole that, while devouring a galaxy, releases a plasma beam at almost the speed of light and it stretches 5,000 light-years.

@LLMJunky Xhigh not available?

And finally, there were some changes to the reasoning levels. For now, both Plan and Code mode are done with Codex 5.2 High. They are not configurable as far as I can tell. Keep in mind, this is a beta 🙃 Surely they will improve features later. Another breakdown loading...

🔥 Codex 0.9.0 is out, and with it, a bunch of new changes. Some of these updates are exactly what you've been waiting for 🫵. Starting with Codex's new Plan Mode! There's been some improvements since my last tweet. This is going to be a long one, so buckle up. 👇

Internet SOFR is emerging: BTC-collateralized stablecoin borrow rate. Interestingly enough, they are almost exactly the same

@dabit3 Writing code… So 2020

I can’t believe we used to actually write code 💀

@shumochu @karpathy You can use AI to replace the teacher asking the questions! So less work

@karpathy We should bring the oral exam back, where a student knowledge is tested by the teacher during a oral defense session. It is more work but worthwhile.

A number of people are talking about implications of AI to schools. I spoke about some of my thoughts to a school board earlier, some highlights: 1. You will never be able to detect the use of AI in homework. Full stop. All "detectors" of AI imo don't really work, can be defeated in various ways, and are in principle doomed to fail. You have to assume that any work done outside classroom has used AI. 2. Therefore, the majority of grading has to shift to in-class work (instead of at-home assignments), in settings where teachers can physically monitor students. The students remain motivated to learn how to solve problems without AI because they know they will be evaluated without it in class later. 3. We want students to be able to use AI, it is here to stay and it is extremely powerful, but we also don't want students to be naked in the world without it. Using the calculator as an example of a historically disruptive technology, school teaches you how to do all the basic math & arithmetic so that you can in principle do it by hand, even if calculators are pervasive and greatly speed up work in practical settings. In addition, you understand what it's doing for you, so should it give you a wrong answer (e.g. you mistyped "prompt"), you should be able to notice it, gut check it, verify it in some other way, etc. The verification ability is especially important in the case of AI, which is presently a lot more fallible in a great variety of ways compared to calculators. 4. A lot of the evaluation settings remain at teacher's discretion and involve a creative design space of no tools, cheatsheets, open book, provided AI responses, direct internet/AI access, etc. TLDR the goal is that the students are proficient in the use of AI, but can also exist without it, and imo the only way to get there is to flip classes around and move the majority of testing to in class settings.

The AI Winter Bull Case for Bitcoin: 1 Private Equity spends billions building massive data centers 2 AI demand cools; the bubble pops 3 Miners buy the distressed facilities for pennies on the dollar Result: Wall St. unwittingly subsidizes the next era of Bitcoin hashrate

@samczsun i.e. You don’t need one single risk model, but the system so people can bet / pay the insurance premium in a capital efficient way Big numbers will figure it out

@samczsun Not an expert in cat bonds either but I believe in capital markets and letting them figuring it out It’s probably possible to “model” risk with history, other security measures… It would only work with long track record, big honeypots, but that’s probably a good 80-20

so how do we secure protocols after they earn "survived the test of time" status? seems like - bug bounty doesnt justify low ev - not worth paying for new audits if its clean - impossible to insure because no way to estimate risk are we cooked or am i missing something (please)

@samczsun To clarify: accepting a certain loss in advance is useful to ensure white hats can get legal and certain access to the typical “keep 10% and we won’t sue”

@samczsun Agree re Seal enforcing accountability Re TVL != treasury, 2 options; - this insurance might look like real world catastrophe insurance? - as a user, at some point you are fine accepting beforehand you can lose a % of the principal for the peace of mind of never losing 100%

@MaximeMB_ @bneiluj @levelsio air you are blasting is colder

@bneiluj @levelsio If it's outdoor I don't see the benefit of an AC vs a fan

The dream of people like @levelsio ….blasting AC in the streets to cool down entire blocks. We went from rich history with lush oases and tree-filled riads to concrete jungles cooled by machines.

@Zup88 @StarkWareLtd You don’t need them !

@jmonteer23 @StarkWareLtd Great, I’d support that too. But the reality is most people in the Bitcoin community don’t even want to think about BTCfi or anything like it. They’re perfectly happy treating BTC purely as a store of value and nothing more.

@StarkWareLtd should in my opinion be laser-focused on 3 things right now: – Privacy features – DeFi – Most importantly: actively pursuing strategic partnerships with TradFi players Why current priorites don’t make sense: Gaming: On-chain gaming shouldn’t be a priority. It’s trivial to bolt crypto features onto in-game economies, that’s where it makes sense. Otherwise, it’s a nice distraction but provides little to no added value. BTC: Nobody I know would ever bridge BTC to BitVM (eg ColliderVM). Op_cat has slim odds of becoming reality anytime soon. Either way, StarkWare can’t drive bitcoin mentality shift alone and there’s zero plan for changing the average bitcoiner mindset. So why pour resources into a direction with no path to adoption? @EliBenSasson

@smiliebg @mfts0 A matter of when (and how frequently) you predetermine it. Traditional buses do it once and maybe review every some years This could be done daily, hourly, or in real-time (other customers are going this way, you are in the way, walk 5 mins and hop in)

@jmonteer23 @mfts0 How exactly is the predetermined route dynamic?

Uber just invented … … the bus

@angeris @fede_intern That’s what I thought. When the tech is finally ready (soon), it will be mainly a “political” issue

@jmonteer23 @fede_intern i think they’re still a chain of blocks, not sure on the former: it will depend on just how much the teams are willing to throw out previous architectural decisions or maintain backwards support

so, uh have we been building blockchains wrong, given what we know today?

@angeris @fede_intern I read it! If the answer is there, it might be too subtle for me… I guess my question is: what needs to happen to turn Ethereum / Bitcoin into the 1GB per block, fully-verifiable-on-your-phone, private, and fast blockchains? Bonus: would they still be called blockchains?

@jmonteer23 @fede_intern idk reading the post might help answer that !

1/5 Bridge is introducing USDB—a programmable, secure stablecoin designed for developers. It’s a new way to embed digital dollars into your app, with rewards built in.