HNHacker News
TopNewBestAskShowJobs

plopilop

591 karma · joined June 2, 2016

submissionscomments
plopilop··on Homomorphically Encrypting CRDTs
As the article mentions, fully homomorphic encryption is insanely slow and inefficient. But I have to say that it is a relatively new field (the first FHE scheme was discovered in 2009), and that the field has immensely progressed over the last decade and a half.

The first FHE scheme required keys of several TB/PB, bootstrapping (an operation that is pivotal in FHE schemes, when too many multiplications are computed) would take thousands of hours. We are now down to keys of "only" 30 MB, and bootstrapping in less than 0.1 second.

Hopefully progress will continue and FHE will become more practical.

plopilop··on What does the end of mathematics look like?
How is that any different than people programming for fun?
plopilop··on Show HN: I created a language called AntiLang – breaking all the conventions
The hegemony of software only accepting . has de facto pushed the standard everywhere for computers, but here in France I still write with a comma, but type with a dot.

A few years ago Excel and some other softwares started to be locale dependent and I never wanted to burn my computer this much

plopilop··on Anthropic: "Applicants should not use AI assistants"
> When have gun laws ever stopped a shooting?

Have you heard of Australia? https://www.sydney.edu.au/news-opinion/news/2018/03/13/gun-l...

plopilop··on Physical Intelligence's first generalist policy AI can finally do your laundry
> The cost of rent is 30€ per square meter per month in Paris.

Only for big enough apartments. 20m² appartements do not go for under 850€, ie around 42€/m². Granted these flats are too small to even have a laundry machine, let alone a laundry folding robot.

plopilop··on Apple Passwords’ Generated Strong Password Format
Switching one uppercase letter in an otherwise lowercase password adds exactly 1 bit of entropy (per switched letter).

On the other hand, any extra lowercase letter will increase the entropy by 4.7 bits (assuming a password on [a-z]).

Given that most passwords have at best 2 uppercase letters, I would argue it is safer to force longer passwords than passwords with at least one uppercase letter.

plopilop··on Programming Zero Knowledge Proofs: From Zero to Hero
You don't necessary get 100% certainty but the probability of success increases exponentially with each new run.

Thus you can get very fast to a probability smaller than you quantum tunnelling through a wall

plopilop··on Microsoft to delay release of Recall AI feature on security concerns
How is this objectively a great feature? This is a spyware that stores screenshots unencrypted (and thus accessible to any other spyware). I am also not convinced that the AI tools would have been offline, thus effectively sharing your whole data with Microsoft (even more than before).

From a privacy perspective, this feature is an abomination

plopilop··on Why can't my mom email me?
Related research:

* Why Johnny can't encrypt (1995): https://people.eecs.berkeley.edu/~tygar/papers/Why_Johnny_Ca...

* Why Johnny still can't encrypt (2006): https://cups.cs.cmu.edu/soups/2006/posters/sheng-poster_abst...

* Why Johnny still, still can't encrypt (2016): https://arxiv.org/pdf/1510.08555.pdf

At this point I really wonder if e-mail is the best solution for encrypted asynchronous communication. E2E systems like Signal or Whatsapp offer a very functional, intuitive way to protect your texts.

plopilop··on Spain expels two US spies for infiltrating secret service
I mean you seem to be simply reducing spying to the GPD, when it varies very much with the policies of a country. Israel is 27th by GDP but is a world leader in espionage, and especially in surveillance tools. Similarly France is good enough at spying that it got offered to join the 5 eyes (it did not work out because France asked the mutual non-spying clause to be applied, which the US refused).

So while it is probable that the US has a bigger spying industry, it's also hasty to dismiss another player by saying "lol 10% of GDP"

> Hermes, Dior, LVMH, etc. I don't think these exactly help spying

Well France carried spying activities in Syria through Lafarge. Sometimes all you need is a foothold into the good places (Syria, or, for L'Oréal, rich circles)

plopilop··on Binance lays off over 1k Employees
The convertibility between money and gold has stopped in the 70s.

Bitcoin allows for instant money transfer, and so does Visa.

The fact that bitcoin has no collateral is a whole other topic.

plopilop··on Binance lays off over 1k Employees
If you really believe that something that is 1 million times less efficient than Visa/MasterCard for a fraction of the service, and that miners really use residential energy when they are active 24/7 and thus by definition a continuous strain on the grid, I have so many scams^W interesting products to sell you

https://digiconomist.net/bitcoin-energy-consumption

plopilop··on Mathematicians discover shape that can tile a wall and never repeat
I meant "a given orientation of the tile will be present many times (or infinitely many)".

It's very probable (I did not read the paper, only their website page) that the tile only occupies a finite amount of orientations in the tiling and therefore at least (and probably more if not all) one orientation will also be present an infinite amount of times.

However this does not imply periodicity of the tiling.

plopilop··on Mathematicians discover shape that can tile a wall and never repeat
The key point (often missed by these articles) is that aperiodic tilings do not have a (infinite) periodic pattern. This means that you cannot draw a shape on these tiles and say: "the tiling is based on infinite repetitions of this shape, and only this shape".

Of course individual tiles will repeat, but never in an infinite periodic pattern.

Edit: a novelty of this paper is that their shape is "truly" aperiodic, which means no matter how hard you try, you will end up with aperiodic tiling. Existing one-shape aperiodic tilings had to add constraints on how to put two shapes next to each other to ensure aperiodicity.

plopilop··on The Point of the Banach-Tarski Theorem
Usually definitions are considered to be finite. Otherwise yes you can just name a number by enumerating its decimals but what's the point.

Edit : actually does not even work, courtesy of the Chaitin's numbers

plopilop··on UK high street banks are quarantining from crypto
I am not pretending that banks are perfect. But regulations are made so these situations happen as little as possible. In cryptocurrencies, such frauds and gross mismanagement are called an ordinary Monday.
plopilop··on UK high street banks are quarantining from crypto
Ah, yes, the freedom to invest all your life savings in obvious unregulated scams such as FTX.

Cryptocurrencies advocate for a very libertarian point of view: you are the only responsible of your money, and if you get scammed, sucks to be you. Banks on the other hand are much more "social": they are protecting you against scammers, and sometimes against yourself.

I know a case of a person falling to a "Hello we're the FBI we need to get your money for an undercover operation but we'll give it back", wiring tens of thousands to scammers. The bank did everything they could, including "losing" the wire order until that person family and friends managed to bring the victim back to their senses. The same story would have been very different with Bitcoin.

People here on HN are tech savvy, and smart enough to not fall into these scams. But advocating for a system that will leave all the most vulnerable members of our society penniless is quite astounding to me. "Some of you may die, but that's a sacrifice I'm willing to make".

plopilop··on Ssss: Shamir's Secret Sharing Scheme (2006)
Sure.

One example is as follows: each party i generates a random polynomial P_i, and n secret shares of that polynomial (j, P_i(j)) for j in 1..n

Then, party i sends (j, P_i(j)) to party j. Party i similarly receives shares (i, P_j(i)) for j=1..n. Party i stores the share (i, sum(P_j(i))).

Then, parties reveal their shares as usual, the secret is then the sum_i P_i(0).

Of course, if the parties are dishonest, you might want some additional safety mechanisms, which can be dealt with with Kate commitments.

This papers https://eprint.iacr.org/2020/504.pdf goes into details, and much more.

plopilop··on Crazy crypto mining story from 2013
> You lose the utility of network effects, etc

Oh yes, you lose a lot of functionalities, but it's not the focus of the computation, simply to estimate, with the current technology, how much it compares to Visa in terms of electricity consumption. To be honest it would maybe make more sense to actually scale Visa down to bitcoin transaction rate, as we can assume more linearity there. The end result is the same though.

> Its like comparing energy costs of bicycles and ocean freighters, but no amount of bicycles could ever tow freight across an ocean.

Quite. However it seems to me its seems unavoidable to do so, since energy efficiency is a big topic these days. Also, while nobody pretends to be able to carry a load across oceans with a bike, it is usually advised to commute by bike rather than a freighter.

plopilop··on Crazy crypto mining story from 2013
> because you'd be looking at blocks the size of several terabytes in order to fit in global scale transactions.

My comparison would rely on running several bitcoin networks in parallel. Of course it does not make sense economically speaking, but it allows for a back of the envelope estimate.

> Electricity for mining is also not directly related to transaction throughput

This is true, but when you look at the C02 consumption (which was the topic of discussion when I first posted, before my OP decided to completely change subject), the mining cost is what you look for.

plopilop··on Crazy crypto mining story from 2013
A real, complete comparison is impossible as the two systems are fundamentally different and too complicated. What I like with that comparison is remains very simple. If you'd scale BTC to the amount of tx operated by VISA/MC, you'd need around 50% of the world electricity, which is clearly more than what Visa takes. Also it's a comparison of the fundamental part of the exchange, i.e. the authority checking for the authorization (visa or the miner).

If we incorporate the lightning network, then we must also take into account the electricity to run the servers for the intermediate exchanges etc. On the other hand, one could also lower the fiat electricity cost by factoring in the cash transactions.

TBH i'd be interested to see a reasonable estimation of the BTC+lightning emission costs per transaction, I'll admit I have no idea how much lightning brings the cost down.

plopilop··on Crazy crypto mining story from 2013
> It currently uses .1% of global energy

ie. about between 300 and a million times more energy than VISA/Mastercard when adjusted to the number of transactions. [0]

> It makes use of stranded energy

Miners who mine 24/7, by definition, do not consume stranded energy. There are no numbers whatsoever that can affirm which part of the miners use stranded energy. On the other hand, when one coal mine got flooded in China last year, 33% of the mining power came to a halt. [1]

> off peak energy stabilizing grid production

So much stabilizing it has caused blackouts in several countries [2]. Texas has resorted to pay miners so they don't cause blackouts [3]

> incentivizes green energy build outs

No, it incentivizes cheap electricity sources. Moreover in the topic of climate change the greenest energy is the one you don't consume.

> with the landfill methane caputure business will soon be the only digital asset that is carbon negative

What? Landfill methane business is basically saying "instead of just burning methane, let's burn it and use the heat to power up some BTC mining". At best, that makes the operation carbon neutral.

If you are trolling, I must say, congrats, you triggered me.

[0] https://digiconomist.net/bitcoin-energy-consumption [1] https://fortune.com/2021/04/20/bitcoin-mining-coal-china-env... [2] https://www.bloomberg.com/news/articles/2022-01-25/kazakhsta..., https://www.bbc.com/news/world-europe-59879760, https://www.cnbc.com/2021/05/26/iran-bans-bitcoin-mining-as-... [3] https://fortune.com/2022/07/12/texas-bitcoin-miners-paid-shu...

plopilop··on France looks into increasing uranium enrichment capacity
Classical numbers for gCO2/kWh do take into account uranium extraction, and yet despite relying heavily on coal as you point out, nuclear (6-12 gCO2/kWh [1]) still remains one of the most competitive power sources [0].

I agree with the rest though.

[0] https://www.ipcc.ch/site/assets/uploads/2018/02/ipcc_wg3_ar5..., page 539 [1] https://www.polytechnique-insights.com/en/columns/energy/car...

plopilop··on On Adobe acquiring Figma
Complete noob in business/whatever this is, why doesn't this look good?
plopilop··on The Ethereum merge is done
There has been some work on the topic, for instance https://eprint.iacr.org/2014/765.pdf

But the main issue in provable security is that you're trying to prove real world things with math, and so far we're quite bad at it. The more mathematical the thing you want to prove is, the better.

plopilop··on How Isaac Newton discovered the binomial power series
I clearly despise academics (and the papers they write) who think that everything should look novel by hiding your thought process.

Sadly, especially in fields where conferences are favored over journals, this seems to be common: you only have a single chance of impressing the reviewer, so you lose little by hiding your thought process, and if possible make your result look more complicated that it actually is.

Many papers I read look needlessly complicated, and I cannot help but wonder why the authors chose this style. I've already been rejected because reviewers found my contributions too simple, however I'd say that the majority prefer to highlight the fact that the paper is "easy to understand". (Of course, sometimes you also need to get told that your contributions are minimal).

Sometimes, I wonder if I have this point of view because some academics like to hide their simple result under a mountain of useless rigor, or because I'm not smart enough to comprehend the necessity of the above rigor and I'm just a dumb engineer who chanced into research. I'd say both are true, but usually not at the same time: some papers are clearly hiding a small contributions, while some are just above me.

plopilop··on Post-quantum encryption contender is taken out by single-core PC and 1 hour
Probably very few or none of them, cryptography getting broken is common. There is no hint that this weakness was introduced on purpose.

Now, did the NSA find the attack on their own, kept it secret and hoped for standardization? One can only speculate.

Also, the NSA does not always try to backdoor everything like they did with DUAL_EC_DRBG. For instance, they modified the S-box of DES during the NIST standardization and refused to elaborate on why. The reason was later found, as people invented differential cryptanalysis and found that the NSA-modified S-box was resistant to it, when the original S-box was not: the NSA actually made DES stronger on purpose.

https://arstechnica.com/information-technology/2013/09/the-n...

plopilop··on Post-quantum encryption contender is taken out by single-core PC and 1 hour
I don't get your point. Cryptography (especially new protocols such as SIKE or anything PQ) must undergo public scrutiny for years or decades, and sometimes get destroyed in the process, that's how it works. Coming up with a usable public key scheme not trivially insecure is already incredibly difficult.

Also, while the attack here is devastating there is no blame to put on SIKE authors: the field is immensely vast, complex and abstract, no one can know everything. The attack relies on a somewhat recent result that is definitely not classical crypto textbook.

Furthermore the cryptosystems used by these companies are the ones that have been tested and standardized through the years, not the brand new algorithm designed by cryptographers (as new algorithms must be scrutinized first).

Also, as another commenter suggests cryptographers are not the ones in charge of the actual implementation at their companies.

Finally, you are making a big confusion between cryptography and security.

plopilop··on Sybil attacks on airdrops
Another funny example is the one of Juno: a whale Sybil-ed the airdrop, and gathered 10% of the tokens. Then, the community voted to strip the whale of most of their tokens.

I find the example interesting because it shows that Layer 2 protocols can very much strip you away from your coins, just like banks/governments can.

(https://beincrypto.com/juno-community-votes-to-reject-tokens...)

plopilop··on Nuclear turn green as EU parliament approves new taxonomy
The massive issue I have with these sources is their reliability. Scotland can produce 98% from wind because they can sell the excess or buy when in deficit from their neighbours who modulate their coal/gas plants. If all its neighbours switch to similar methods, I don't see how we can have reliability on a wide scale.

Usual fluctuations (eg. no photovoltaic by night, more wind in the afternoon) can be planned for, but local events such as big clouds or no wind are frequent but have a great impact on

← PreviousPage 2 of 6Next →