HNHacker News
TopNewBestAskShowJobs

dmarchand90

1,343 karma · joined May 16, 2019

submissionscomments
dmarchand90··on Microgravity-induced muscle atrophy: Implications for long-duration spaceflights
I wonder if you could bypass these issues with strength training machines designed for space. E.g a squat machine that pushes down from above, cable tension based arm movements, etc. I bet if you're clever you can probably have a small room that can get the essential exercises in and takes up a similar amount of room as a toilet
dmarchand90··on Fair coins tend to land on the same side they started
Excellent ig nobel prize contender here
dmarchand90··on Human trials of artificial wombs could start soon
Yeah they tried this in the late 19th and early 20th century. The problem (actually very similar to real strongly typed languages) is everything gets very verbose and impractical for ordinary people.

Actually you could argue that we have indeed created a strongly typed dialect of English which is legalese. So a version that is very precise but to complex and verbose for most people and most spheres in life

dmarchand90··on Refact Code LLM: 1.6B LLM for code that reaches 32% HumanEval
Really cool stuff! Do you have any recommendations of where we could learn more?
dmarchand90··on Google was founded 25 years ago today
The nice thing with ChatGPT is they've established a viable funding model through paid subscriptions from day 1.

On the other hand I predict that we will see the next ten years as a 'golden age' of steadily improving performance from ChatGPT, which will then plateau and 'enshittify'.

dmarchand90··on Vale.sh – A Linter for Prose
I think it's the lack of examples. Also, because this is very cutting edge, I'd like to know a bit of the technology behind it to know the strengths and limits. E.g. is it based on hard Coded grammar and thus a bit stiff and dumb? Or is it llm based and therefore with data control issues?

It would also be nice to see examples of what it can do. Again, if it's hard Coded grammar it would be nice to see the complexity it can handle, if llm, the lengths the token limits can manage

dmarchand90··on Ask HN: Why did Python win?
In clean code there is a chapter about eliminating useless comments. I think this has some merits. Think of the following made up example:

sum = sum_two_numbers(number1, number2)

Probably doesn't need a comment.

The other argument is that comments need to be maintained and are subject to decay. E.g. the code they are commenting on changes.

The book goes into other examples but i think the idea isn't to eliminate comments but to be thoughtful and judicious in which you use

dmarchand90··on Ask HN: Why did Python win?
Perl and python have opposite philosophies with regards to standards. Python prefers a standard "pythonic" way, while perl had its "there is more than one way to do it".

It would seem that having a standard is more popular.

dmarchand90··on Show HN: I'm sharing my girlfriend with the world
All of this reminds me of the the old fake infomercial "live forever as you are now" https://youtu.be/xg29TuWo0Yo?si=yvFa-vFIEqD-Swx4
dmarchand90··on I won't deliberately use AI for programming
Oh for sure! But you wouldn't argue to do this exclusively though
dmarchand90··on I won't deliberately use AI for programming
Yeah this logic would have us writing everything in assembly and only using textbooks.
dmarchand90··on Cash payments above €3000 to be outlawed in Netherlands
For me this translation is extremely easy to understand:

What is the maximum amount I can pay in cash? At this time, you may pay for all purchases in cash. For cash purchases above € 10,000, the seller must conduct a customer due diligence. But not all sellers accept cash. The cabinet is working on a bill to ban cash payments from € 3,000.

Pay for purchases in cash At this time, you may pay for all purchases in cash. It doesn't matter how high the amount is. A seller can decide not to accept cash above a certain amount. Or not accepting cash at all .

Customer research for purchases above € 10,000 Do you make a purchase that costs more than € 10,000 and do you pay in cash? Then the seller must conduct a customer due diligence . This check is mandatory to prevent money laundering and terrorist financing.

Bill cash payments up to € 3,000 A bill is being drafted to ban cash payments from €3,000 . This applies to all buyers and sellers of goods. When the new law comes into effect, you may no longer pay for purchases from € 3,000 in cash.

dmarchand90··on Cash payments above €3000 to be outlawed in Netherlands
How important is this rule really? Particularly when the title gives the main message. I ask because auto translation is almost perfect for at least all the European languages. I agree maybe there should be a bias towards English but this seems a reasonable exception here
dmarchand90··on ‘Here Begins the Dark Sea’ Review: A Map from the Future
Works for me
dmarchand90··on Blaise Pascal: The Infinite Spaces, Alienation, and the Wager
My solace lies in the realization that there's an uncountably infinite number of real numbers between zero and one. This means every moment possesses its own unique infinity if we choose to see it.

"If we take eternity to mean not infinite temporal duration but timelessness, then eternal life belongs to those who live in the present. Our life has no end in just the way in which our visual field has no limits" Wittgenstein

dmarchand90··on MetaGPT: Meta Programming for Multi-Agent Collaborative Framework
I suspect the divide may end up similar to that between a start up and a large company. Yes, some problems are better solved by a single person. But others benefit from a division of thought even if there is some bureaucratic overhead.
dmarchand90··on MetaGPT: Meta Programming for Multi-Agent Collaborative Framework
I suspect the real value is more from an architecture and caching perspective.

Architecture as in you may have multiple GPT instances being prompted to look at a problem from different angles: "Analyse the problem as a pessimist", "As an optimist", "as a mathematician", "an engineer", "a philosopher", etc.

Caching, as in determining what you can store from these outputs, e.g. "give a numeric 1-10 score of what you think of this product as a LGBT-friendly conservative from the Midwest", etc.

dmarchand90··on The average Manhattan rent just hit a new record of $5,588 a month
Asking users who live in New York. How much does this matter? Is the commute from Brooklyn or Queens reasonable? If so how are the prices? The Bronx has a bad reputation (as a foreigner) but is that even true?
dmarchand90··on What Is Happening with ChatGPT?
I would guess if you look at all the literature in the world the majority of cases that have a word repeated a large number of times its either a prayer, e.g. catholic Rosary "Hail Mary, full of grace...", or madness, e.g., the typewriter in the shining "all work and no play makes...". So repeating something may, at least occasionally, put the AI in that frame of mind
dmarchand90··on Paul Graham on Ambition, Art, and Evaluating Talent (Ep. 186)
Ugh so embarrassing, thank you
dmarchand90··on Paul Graham on Ambition, Art, and Evaluating Talent (Ep. 186)
Hard to say. The answer has been a bit contaminated since the late 1700's because of Gibbon's attacks when he wrote "Decline and Fall of the Roman Empire". I strongly suspect that the results would be exactly the same, modulo the butterfly effects, meaning everything would be totally different in every possible way.

Actually we have 'zero' clue (and we never will) know how sensitive history is to 'thermal' noise, as we are stuck running N=1 simulation.

dmarchand90··on Paul Graham on Ambition, Art, and Evaluating Talent (Ep. 186)
I found the section where they ask "what would happen if Rome didn't fall" and then they speculate on China for a while to be very annoying.

We know the answer. Because only half the empire fell. Constantinople continued another thousand years after the fall of Rome. So there's your answer, you would get something like Constantinople.

dmarchand90··on LK-99: Phonon bands, Localized Flat Band Magnetism, Models and Chemical Analysis
Damn, looks like it's a no then :(
dmarchand90··on Replace peer review with “peer replication” (2021)
I believe that, contrary to popular belief, the implementation of this system would lead to a substantial increase in productivity in the long run. Here's why:

Currently, a significant proportion of research results in various fields cannot be reproduced. This essentially means that a lot of work turns out to be flawed, leading to wasted efforts (you can refer to the 'reproducibility crisis' for more context). Moreover, future research often builds upon this erroneous information, wasting even more resources. As a result, academic journals get cluttered with substandard work, making them increasingly difficult to monitor and comprehend. Additionally, the overall quality of written communication deteriorates as emphasis shifts from the accurate transfer and reproduction of knowledge to the inflated portrayal of novelty.

Now consider a scenario where 50% of all research is dedicated to reproduction. Although this may seem to decelerate progress in the short term, it ensures a more consistent and reliable advancement in the long term. The quality of writing would likely improve to facilitate replication. Furthermore, research methodology would be disseminated more quickly, enhancing overall research effectiveness.

dmarchand90··on LK-99 is an online sensation but replication efforts fall short
The important thing to understand is that only the scientific publications in Nature matter. These articles are written by world-class scientists and are taken very seriously. In contrast, the journalism section is akin to any random newspaper. It is generally written by standard journalists and is intended for a mass audience.
dmarchand90··on Observation of zero resistance above 100 K in Pb₁₀₋ₓCuₓ(PO₄)₆O
What makes me excited is there is very good theoretical work, using 'standard techniques ' that shows it either is a superconductor or will have very good properties in designing one.

Cold fusion, the EM drive, time traveling neutrinos, and all the other false excitement never had a theoretical backing. So 'guarded optimism' seems warranted

dmarchand90··on Observation of zero resistance above 100 K in Pb₁₀₋ₓCuₓ(PO₄)₆O
The big difference is that standard theoretical techniques have indicated the compound to be promising from an electronic structure point of view and confirmed that it would be tricky to synthesize just right.

Cold fusion always was dependent on new physics to be discovered

dmarchand90··on Unconfirmed video showing potential LK-99 sample exhibiting the Meissner effect
It's a great question. I don't think anyone has a certain answer but it's definitely a real risk. The quick example is graphene which has all kinds of interesting properties, but, as far as I can tell, no practical applications primarily due to production issues
dmarchand90··on Why transformative artificial intelligence is hard to achieve
Yes but that's not the question either. Chatgpt can probably publish in a journal already. The question is whether it can make impactful work. This is very unclear if most people could do even if working on it 100%
dmarchand90··on AI researcher Geoffrey Hinton thinks AI has or will have emotions
I'm still not persuaded that we can really know what's going on "in there". Agreed that it only has an extremely limited short term memory (~2k tokens or whatever) but ultimately what goes on between the weights of the network is a mystery.

I do agree that it is almost certainly not sentient. But, I'm very skeptical to the degree we can be certain that these objects that can pass the Turing test don't have some form of awareness. (Which would probably be something closer to a set of disconnected flashes than our own continuous film stream).

I just think humility and caution is the best policy. I always think of how doctors used to be very confident young babies felt no pain and could be operated on without anesthesia (this text is entirely my own)

← PreviousPage 3 of 12Next →