Hacker Newsnew | past | comments | ask | show | jobs | submit | nikhizzle's commentslogin

I think the interesting thing here for those of us who use open source frameworks is that we can ask the LLM to look at the source to find the answer (eg. Pytorch or Phoenix in my case). For closed source libraries I do not know.

I may eventually get one of these just to use with Claude code. Been looking for the lightest best machine to use with agents.

Maybe dictation is the way to go? It’s a quick way to interact with agents and works really well.

Shellfish on iOS to ssh into a vps with tmux with Gemini-cli, lazygit and neovim worked quite well for me.

The clicks keyboard does not have ctrl, arrows, page up, down or really any special keys so I’m not sure it would be that much more pleasant. I know iOS keyboard has been quite meh in the recent releases but for thumb typing I’m not convinced that physical keyboard are superior.


sample output here - https://www.science.org/action/downloadSupplement?doi=10.112...

This is at the very beginning of begin feasible. I do not know anything about photonics, maybe someone who does can comment on scalability?


The raw materials: diffractive optical elements and single mode fibers from a materials perspective are all quite easy to manufacture. The primarily limitation with miniaturization is the single-mode fibers, which are limited by the optical wavelength you are using and the index of the fiber. For a conventional silica optical fiber, this is probably around ~100 nm diameter at a minimum. Newer materials can definitely change this 2-3x, but I'm not aware of anything more fundamental.

So in general this would be something that you would potentially be able to see in cars, but unlikely consumer electronics or handhelds without a modification in the operational principle (eg time-multiplexing to reduce the required number of fibers).

My personal opinion is that competing on low-power and small-scale is a lost cause for photonic computing. In terms of absolute energy efficiency and absolute miniaturization, photonics will never win. But at larger energy scales and larger systems, photonics can reach a regime where higher parallel throughput will dominate.


Thanks for sharing this. Needed something meaningful this morning.


You're welcome. He also remained in contact with his crew until their deaths.


Armchair economist here - one implication of this is that a crypto liquidation will cause global interest rates to spike at a time when they will need to be lower to calm the markets.

Selling massive amounts of debt with no additional demand means the required return must be higher.


Stablecoins are now a regulated industry with laws in place to address this very concern. Tether is currently not compliant, but as the largest player in the space, has a great incentive to maintain its dominance.


Regulated industry.. like banking, stocks, and mortgage origination!


What would the regulation ... enforce?

Some sort of behavior by Tether and others?


Regulators also have enforcement functions.

Are you asking how laws work?


I’m just curious what is being enforced to prevent what exactly.

I think I’m missing what scenario people are thinking of in general.


I'm not sure I see that, depending what you mean by crypto liquidation. If you mean the prices of dogecoin etc. falling then that mostly just effects the number of Tethers changing hands between one speculator and another and wouldn't really affect the bond market.

If you mean Tether holders redeeming them for US dollars, that would involve selling treasuries but I doubt it would drive the price down that much. That's a very liquid market.


I mean tether redemptions. Hedge funds liquidating treasuries last year significantly distorted yields for a short while.


Good point but it seems like Tether is very healthy, it has the double of daily volume than Bitcoin now and doesn't really depend on crypto rails [1].

[1] Stable Coins ⊄ Blockchains: https://www.linkedin.com/pulse/stable-coins-blockchains-seba...


that's very armchair indeed

a crypto liquidation results in more people going to Tether. "tethered" and "tethering" has been a verb in the crypto space for like 10 years

when market demand of tether is too great the value goes about $1.00 and the organization relies on arbitragers to deposit more to cause the minting of Tethers at $1.00 and selling it into the open market if the price is above $1.00 pushing the market rate back down to $1.00

Tethers in existence still continue to grow in that scenario

Its nearly 1:1 backing, most of the time, even a mass redemption event of Tethers will be orderly and fine. those crisis of confidence have alreay occurred, those stress tests have already happened, far faaaar beyond what any bank would survive

all fiat collateralized stablecoins function the same way and there are many case studies, actual events that happened, that show it occurring orderly, uneventfully.


> when market demand of tether is too great the value goes about $1.00 and the organization relies on arbitragers to deposit more to cause the minting of Tethers at $1.00 and selling it into the open market if the price is above $1.00 pushing the market rate back down to $1.00

I'm not sure of this. USDT can only be minted by tether [0] (line 406), hopefully when they acquire more treasuries.

I am not exactly sure where the price of USDT come from and I am pretty sure there is not only one. But I would guess it is an aggregate of the prices on exchanges.

The chainlink oracle [1] is probably the most authoritative one.

- [0] https://etherscan.io/token/0xdac17f958d2ee523a2206206994597c...

- [1] https://data.chain.link/feeds/ethereum/mainnet/usdt-usd


Nobody is “setting the price”, people trade. supply and demand. on many venues and contracts. the oracles just read from those (and sometimes the venues and contracts read from the oracles)

when tethers and trading at $1.04 or anything higher than $1.00

arbitrageurs wire money to their account at the Tether organization (bitfinex, otc services) to instruct the organization to mint Tether

the organization does everything necessary: takes the deposited money and buys US treasuries, mints the equivalent amount of Tethers and gives those tethers to the customer

The customer deposited $1 and received 1 tether. Some exchange venues have people rushing to buy 1 Tether for $1.04

Customer sells their tether to them and has some other form of crypto that they can try to get back into dollars and do it all over again, until flooding the market with Tether supply back to $1.00

this happens all day every day for a decade, more pronounced during panic selling periods


The USA is going to have a fantastic fiscal reputation if/when that has happened.


deleted


The US Treasury market is BY FAR the deepest most liquid market in the world.


The US housing market is the largest market in the world.


Thanks for this, this is very useful for a related ai project I’m working on.


A job feed for remote jobs - https://tangerinefeed.net/

This is something I’ve needed myself over the last few years as jobs become shorter and shorter lived. Keep on improving it as some kind of compulsion.


Looks good! Seems to not be bringing in the requirements section of the JDs?


Thanks! Will take a look.


Ex-Apple File System engineer here who shared an office with the other ZFS lead at the time. Can confirm they link to the wrong profile for Don Brady.

This is the correct person: https://github.com/don-brady

Also can confirm Don is one of the kindest, nicest principal engineer level people I’ve worked with in my career. Always had time to mentor and assist.


Not sure how I fat-fingered Don's LinkedIn, but I'm updating that 9-year-old typo. Agreed that Don is a delight. In the years after this article I got to collaborate more with him, but left Delphix before he joined to work on ZFS.


Given your expertise, any chance you can comment on the risk of data corruption on APFS given that it only checksums metadata?


I moved out of the kernel in 2008 and never went back, so don’t have a wise opinion here which would be current.


I have a rough idea of my friend’s schedules, and I call them when I’m driving the kids around, or we text. Staying in constant contact with a few very close friends about my mental health has floated me through my mid-40s.

A tip I got from a friend in his 60s was that even when you lose friends, life is great because you constantly have the opportunity to find new ones. I am in a new close friends renaissance in my 40s, just be vulnerable and don’t take rejection personally.


I'd definitely be better off if I didn't take rejection personally

But it's hard for me to draw the line between being vulnerable and say, oversharing or dumping


So which one is it? Code is contract and he should get to keep the money. Or crypto is governed by laws outside of crypto and so he violated the “spirit” of the code and hence is a criminal?

It seems like right now the crypto industry makes the decision to their convenience on a daily basis.


Purity goes out the window when there's real money involved. And means that in cryptocurrency, you only own what the government grants that you own.

It'll be interesting how this gets resolved by Canadian courts.

And this is rich: “A bad actor not brought to justice and held to account for one act of fraud will surely commit another”


> It'll be interesting how this gets resolved by Canadian courts.

He's being prosecuted by the US DOJ, EDNY. Not Canada.


[flagged]


This comment adds no value to the discussion and tries to politicize the conversation.


Code is contract and disputes are handled by the courts. There's no such thing as a purely extrajudicial contract, is there?


I think the name smart contract is misleading because if you believe that the code is law then there is no actual contract in the smart contract. There is no meeting of the minds, no agreement on what the contract means or what is considered customary. Just a machine floating in the ETHer you can interact with. You owe the machine nothing and it owes you nothing.


There is a third way: Private adjudication. I see no reason why the crypto community couldn't run its own private court system, similar to what e.g. Randy Barnett describes in chapter 5 of Anarchy and the Law.

(Obligatory, I don't work in crypto or have any special connection to it, I just think people forget this third option even exists when it's really the most common way most private industries actually resolve disputes most of the time.)


It already does run its own private court system, it's the consensus process that is used to upgrade the protocol. If there was enough consensus there could be a fork that returned the funds, but that will never happen.


Sure, but suppose people decide that's not good enough.


what "people"?


simple, those who lose to the system they’ve supported until it bleed up in their face


Private court system? How does it work? Regular court system works because the government has the threat of violence, has the muscle and rights to hunt, capture and detain. if "private court" rules against me and I refuse to obey, how will they make me obey? There is no "or else" embedded in it so it will be useless.


its a cryptocurrency, the courts legitimacy would likely come from the community forking as necessary to resolve issues as identified by the court, in the same way the us judicial branch gets its legitimacy from the us executive branch employing force as necessary to resolve issues as identified by that court.

its a broader group to convince, sure, but there is a clear 'or else' from which the court can get that legitimacy. 'return the money or else we will return it for you' is a meaningful or else


There's a service called Kleros that does exactly this


There definitely some hypocrisy, but it might work differently in the law.

As devs, we might claim that 'code is the law' but my guess is that the law does not care. That is, one cannot overwrite property laws by a few lines of code.

Consider how disclaimers work--we are increasingly putting limitations on what rights you can contractually forfeit.

This will be interting to watch.


It seems absolutely bonkers to me that someone would write a smart contract that lets them bleed $50m without automatically stopping after they lose the first $1/10/100k.


That seems hard to implement. Presumably the move was all based on regular transactions. You can't just say, "If we're losing money, stop it", you have to concretely specify the conditions.


In the real world locks are meant to keep honest people honest and slow down the dishonest people until someone notices and stops them.

There’s a world where crypto could be sold the same way, but the sycophants drowned that out for long enough that we aren’t in the Trough Disillusionment now so much as the Trough of Open Mockery.


We have laws, yes.


Yes, we have laws. When should the law intrude on the private transaction of two parties? Typically, the law holds both parties to their contractual agreement. If those two parties have contracted to abide by the output of an algorithm, can the law distinguish good faith manipulation of algorithmic inputs to benefit oneself from bad faith manipulation of algorithmic inputs to benefit oneself? Given that the whole point of a smart contract is to encode the terms of the agreement as code, when is it appropriate to step in and alter that agreement?


See also: tether.


Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: