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

Sometimes I wish some kind of weird disaster would strike that somehow only erases the protocols and styling/markup languages invented in the last 60 years -- without losing any data -- to force us to start over, but with the benefit of hindsight.

Oh, and JavaScript.


This reminded of a CK Lewis bit about how modern humans deploy a lot of resources trying to save "weak" babies, and thus undoing evolution's natural selection process.


That perspective is always such a fundamental misunderstanding of evolution and natural selection. (Yes I know It's for a comedy bit but I see this way too often).

Darwin's theory of evolution and natural selection were never really about directly competing against other members of their species. There was certainly a component of that but natural selection is predominantly about competing against nature itself.

It's all about developing traits that help a given individual or community/ecosystem survive and thrive. And unsurprisingly in most ecosystems it's not competition from peers but rather competing against weather, environmental conditions, and the food chain/predators. So what you see is that at basically every single level (from plants and microbes, up through insects, birds, mammals, and at all stages of human history) you have a constant push for mutualistic behaviors.

It's why birds warn their entire ecosystem (including other bird species and non-bird species) about predators and danger. Or as another bird example, migratory birds will cooperate and share food even when migrating with birds of different species. Anything that can bolster the ability to survive and thrive for the community as a whole (and often entire ecosystem) ends up driving evolution far more than advantages for a single individual. Doubly so with punishing adversarial advantages for individual that end up disproportionately harming the community/whole.


That's only part of the truth. Animals do cooperate within and even across species, but they also compete, even within a species - wolves, ants, and chimpanzees are all territorial (as are many others), and the latter two are known to engage in war within their own species: https://www.livescience.com/animals/land-mammals/a-decade-lo...

And the competing against nature itself you mention, is often determined by the territory a group is able to claim. Some places get drought, others freeze, and in others food is plentiful. Nature may not be a free-for-all deathmatch, but it's not a pacifist coop either. At least, most species don't behave that way.


Oh certainly. But that's the thing. Even with species being territorial, that serves a broader purpose in the ecosystem. Territoriality for predators is important to prevent concentration of predators, overpredation, and then depletion of prey species (which has many downstream effects).

And because of that, territoriality tends to be fairly low in most species until the food supply becomes constrained. And even then it's a gradient where hostilities generally only escalate out of desperation rather than innate competition. i.e. Competing between individuals or communities tends to occur mainly when they fail to compete against the environment and run out of other options.

But really my point was just about the general sentiment that it's "against evolution" or "against natural selection" to help the weak and that doing so is something that humans do out of a unique sense of love or kindness or whatever.


If natural selection is about avoiding death, then nature must be doing a poor job since everything is dying in the end.

If killing the unfit is the way to go, you should kill your babies until they become immortal.

Natural selection has always been about reproduction.


So...Eugenics, then?


Already happening at the in vitro level, might be possible in vivo as well. Neither require the more abusive approaches from the first eugenics era.


Yep. It's inevitable and societies will have to grapple with it far sooner than most thing.


Eugenics is also "undoing evolution's natural selection process".


In todays world, yes, that is back on the table.


As long as some people and societies have more children than others, evolution continues.


You're talking about physical weakness which can be caused by non-genetic factors. Such a person may turn out to have a great intellect or other personal quality.

However, the big story in the west is that most sexual congress does not produce babies anymore.


I don't use TikTok but spend some time on Instagram. Despite the format, I enjoy a lot of intellectually stimulating content (and, sometimes, conversations) on the platform.

Sometimes a friend would show me their feed and I'd be shocked at how different the content they are presented by their version of the algorithm.

There are a lot of people putting a lot of effort to create very interesting content and we should not belittle their work just to fein intellectual superiority.

There's really nothing inherently wrong about the format.


> There's really nothing inherently wrong about the format.

"Our conventional response to all media, namely that it is how they are used that counts, is the numb stance of the technological idiot. For the 'content' of a medium is like the juicy piece of meat carried by the burglar to distract the watchdog of the mind...The effects of technology do not occur at the level of opinions or concepts, but alter sense ratios or patterns of perception steadily and without any resistance. The serious artist is the only person able to encounter technology with impunity, just because he is an expert aware of the changes in sense perception." — Marshal McLuhan, Understanding Media


I feel like McLuhan is so thoroughly accepted as gospel now that it’s refreshing to see someone casually dismiss the idea out of hand like OC. I would love to see a serious exploration of the argument against McLuhan in 2025, just for fun.


You mean aside from the fact that his quasi-mystical, self-important proclamations have no factual basis? I personally dislike how his ideas cast a cynical veneer over truly transformative changes without actually contributing anything useful to the conversation.

Who takes this stuff as gospel? Academics who study critical theory?


McLuhan is anything but a cynic. It sounds like you haven’t read him.


I suspect he got right more than he got wrong.

Who is OC?


"Original Commenter"

I have not seen that particular jargon before but it rhymes with OP (Original Poster) that I was able to figure it out from context.


There is 100% some egotistic intellectual superiority as someone who consumes long form written content and long form video content. But its hard not to look down on short form video as a format and the culture of the medium. It seems clear to me that shorter form content is forced to "simplify" topics and sweeten narratives to retain the viewers attention. I've seen a ton of short form videos where I came away thinking wow what a great little video i really feel like i've learnt something. But then I double check to see if its actually true and its simplified to the point of being misleading. My sister shows me these videos all the time and its misleading or completely untrue majority of the time.


I don't think the article wants to belittle their work. The format isn't wrong, but the way the content is structured is troubling; it is to cater to the algorithm by tapping into the parts of our brain that are not meant to be stimulated all the time.

Any form of media isn't bad as long as it's analytical, i.e., one that forces you not just to be an observer all the time. Most content on such platforms is designed to keep your brain constantly stimulated so you never shift your attention to think. The stimulation and the dopamine hits just keep you hooked to it. They give little time for contemplation, encouraging passive consumption.

When consuming long‑form content in any format, you get bored or drift just enough to think your own thoughts as you consume. But when consuming short‑form content, you are forced not to think unless you choose to pause; if you get distracted, you might think you missed something, which you don't want to do.

Information‑dense content is not good in any way, whether academic or entertainment. It doesn't leave you with any time to think on your own, discuss with yourself or the creator, dismiss some faulty thoughts, and eventually form an opinion of your own that you want to discuss with someone, somewhere.

That being said, not everything has to be long form content. Short content can provide concise information where needed, also serve as a gateway to deeper exploration, if the viewer follows up. I am not sure how that can be encouraged as most do not choose to do, as they are drowned with it and never get time to explore deeper into topics they want to.


I recently read "Chip War" and it talked about an era (around the 80s and 90s) were american dominance on electronics (and economy) seemed in deep decline.

Japan was the next big thing.

But the collective efforts of some government agencies, academia and the private sector helped reverse the trend.

American dominance is sure not a given but with an almost century of inertia, all hope is not lost (especially compared to the alternative).


> But the collective efforts of some government agencies, academia and the private sector helped reverse the trend.

Well that's the key. The current administration is doing its best to sabotage science.


I get it. But what I'm saying is that the impact of a single misguided administration, while can be very devastating, is not enough to write off american super power status in research.

With appropriate planning and funding, the next administration can definitely reverse the trend.


The current administration is braking hard against the inertia.


Even the concerted effort of a competent administration wouldn't be enough to cancel a system that's a century in the making.

Keyword: competent.


The problem is that not all governments constrain themselves to only use requests and/or allow companies to challenge said requests. And that's at least what you get dealing with Apple and the USA (most of the time).


(Most of the time) is as true as you trust the federal government that demonstrably lies and “loses” records whenever it feels is convenient for them.


This is a fairly new question; from the early 20th century, iirc.

There were many questions with no answers for literal centuries and thousands trying, and failing, to crack them. A solution was ultimately found despite that.

A new "math" might be needed, but an answer (affirming or not) will be found.


It is fairly new, but very relevant for daily life, like many others are not. Thousands of people have tried to write smart algorithms to solve NP problems and many have thought they found an algorithm in P only to be disproven later.

Whether the Riemann hypotesis is true or not, is not going to have any practical effect, accept for a small group of mathematisians who are working on it. Most people do not know what a Field medal is nor care about it.


> A new "math" might be needed, but an answer (affirming or not) will be found.

What if there exists a proof that P!=NP, but the shortest possible proof of that proposition is a googolplex symbols that long? Then P!=NP would be true, and provable and knowable in theory, yet eternally unprovable and unknowable in practice


That's exactly the kind of situation I had in mind when I wrote that.

Goodstein’s theory would take more symbols than there are atoms in the observable universe to write down in "classic" maths. To "fix" this, mathematicians had to use a "new" way of thinking about infinity known as transfinite induction.

I think if we're smart enough to detect(?) a proof, we'll find a way to express it in a finite manner.


Apparently there are multiple studies that show a link between listening to heavy metal and being 'happier' and/or 'less angry':

  - What Makes Metalheads Happy? A Phenomenological Analysis of Flow Experiences in Metal Musicians [0]
  - Extreme metal music and anger processing [1]
  - The effects of heavy metal music on arousal and anger [2]
[0]: https://www.researchgate.net/publication/321893408_What_Make...

[1]: https://www.frontiersin.org/journals/human-neuroscience/arti...

[2]: https://psycnet.apa.org/record/1997-05014-002


I suspect that coming up with a set of rules as to who would and wouldn't be honored in such a manner would inevitably lead to some real-world violence at some point in the future.

Let's leave this out of our commonly used code.


Yeah. The new challenge seems easier to solve since it basically is hand-holding the LLMs into what the result should look like.

I think a more challenging, well, challenge, would be to offer an even more absurd scenario and see how the model handles it.

Example: generate an svg of a pelican and a mongoose eating popcorn inside a pyramid-shaped vehicle flying around Jupiter. Result: https://imgur.com/a/TBGYChc


I like the hand-holding because it's a better test of how well models can follow more detailed instructions.

I was inspired by Max Woolf's nano banana test prompts: https://minimaxir.com/2025/11/nano-banana-prompts/


That's a valid point but I'd argue the new test would be then interesting to couple with the original one, not to replace it.

Do you think it would be reasonable to include both in future reviews, at least for the sake of back-compatibility (and comparability)?


Yeah I'm going to keep on using the old one as well.


Which model did you use in the example result? It looks fantastic.


I don't think it would be a good idea to publish it on a prime source of training data.


He could post an encrypted version and post the key with it to avoid it being trained on?


What makes you think it wouldn't end up in the training set anyway?


I wouldn't underestimate the intelligence of agentic AI, despite how stupid they are today.


Every AI corp has people reading HN.


This sounds like paranoia to me to be honest. Please tell me I'm wrong.

I could have easily come up with just the same claim, without seeing the benchmark, it doesn't exist.

Maybe if we weren't anonymous and your profile leads to credentials that you have experience in this field, otherwise I don't believe it without seeing/testing myself.


but they've asked all the AI models this question. Whatever you tell an AI model is also in its training data


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

Search: