The story was flagged by many users. The problem with articles like this is the discussions are repetitive and predictable. We rarely see anyone approaching them with genuine curiosity. The topic of whether this president and administration are befitting of particular labels and historical analogies has been continually discussed (in broader society and on HN) since about 2015. And in the discussions we generally just see people trying to justify why they believe what they already believed about the topic, sometimes quite belligerently.
This is why discussions about politics are generally bad on online forums (and considered to be best avoided at dinner parties); it’s a domain in which people’s belief about the topic is deeply entangled with their identity, and by definition, people get defensive and hostile when their identity is thrown into question. Thus, they work much harder to justify why they were already right about the topic, instead of seeking to learn anything new.
The kind of politics discussion that would be good to see much more of on HN would explore the question: if we were to agree that the state of politics globally is terrible (I certainly do), what actions can ordinary people like us working in technology do to make things better?
The "discussions" are repetitive because the sparks of curious and constructive discussion are shouted down or used up by people who just want to rail against the framing. If you want to support intellectually curious discussion, then letting whole topics get shut down because they are partisan-inflammatory is the wrong thing to do - this merely supports the dynamic of the hecklers' veto. Rather you need to police the partisan disruptors better.
There is a reason there is no downvoting on submissions, right? And yet we've now gotten to a place where submission flags have become an effective downvote. I don't know if this particular post would have generated thoughtful discussion, but I do know that only seeing this submission by chance a day later all of the interest has passed it by.
Comments, when they say "flagged", are also dead. But articles say "flagged" for some number of flags, but they aren't dead yet. It takes more flags to kill them. "Flagged" is just a notification that an article has received some flags, but it doesn't actually change anything yet.
I have not found any good places. (Perhaps that's the point though.)
Politics every day, sadly, says something about the times we're living in. We probably could have blissfully ignored politics in the early days of the internet—hopefully can again some day…
Except when it is political content against the EU or some European nation doing something Americans don't like. Then it is very interesting to HN, never flagged and raised to the front page.
>Otherwise it quickly becomes politics all day everyday.
No, it doesn't. This is one of Hacker News' weird phobias but it doesn't reflect reality. I know the mods believe it too so there's no point in debating it but even Reddit isn't politics all day every day. The nature of the community here is a self-correcting mechanism. This thread is not a flamewar, the posted article isn't low quality (certainly not on a forum where posting Twitter posts and Wikipedia articles is allowed,) and it poses literally no threat to the community, but HN still treats it like a cancer.
> The nature of the community here is a self-correcting mechanism
As if by magic ;)
> This thread is not a flamewar, the posted article isn't low quality... it poses literally no threat to the community, but HN still treats it like a cancer
These threads are bad for HN because they’re not new topics and they don’t yield new insights. The question of whether this label applies to this administration has been discussed for over a decade. What difference does it make to talk about it again? Who is going to be changing their mind or their plans now, in response to this post? Do we ever see any interesting new ideas in these threads?
If we had a steady supply of articles that spawned discussions on HN that could generate new ideas about how to address the political/economic dysfunction we see all over the world, we’d happily have them on the front page every day.
The flagging mechanism, as it exists, feels a little… cowardly (?) to me.
What if, instead of the existing mechanism, we posted in the comments, "Please Flag", and follow that with our rationale.
Mods could "read the room" (so to speak) and flag the article. You might still argue as to whether an article should have been flagged but at least the receipts exist to show why the mods acted.
America does have basically every characteristic of fascism on every important list of fascism characteristics ever made.
That's actually kind of important to the tech community, considering we are wildly complicit in this.
So, maybe consider that more than "politics junkies" might be interested in this, and that the tech billionaires might have a vested interest in making sure stories like this get flagged (very easily done).
> the tech billionaires might have a vested interest in making sure stories like this get flagged
Interestingly, this "anyone with an opinion different from mine must be a paid shill" argument doesn't pop quite as often in the discussions about Clovis Culture tools, Roman Empire letters, or pre-Linotype typesetting -- the fact that makes me think that maybe keeping politics out of HN is actually a good thing.
I don't live in the US so this doesn't quite enrage me, especially since I'm aware that every US president gets tens (hundreds?) of millions of dollars in donations to their election campaigns, so a gold statue does not look like a big deal when compared to that.
> "anyone with an opinion different from mine must be a paid shill"
Not remotely what I said. One of the better HN guidelines here is to try and interpret comments you read in the best possible light. I recommend it.
What I actually said is simply correct - there are tech billionaires who do have strong reasons to flag certain topics on HN.
And there's no way to stop them from doing so. We rely solely on 100% opaque moderation to unflag stories.
There's no shortage of people who have complained about how often threads concerning Musk, DOGE, the Lawnmower guy, Thiel, certain genocidal countries etc get wiped from here...
Do you see those threads? Unless you have [showdead] on, and browse /active, almost certainly not... Because they get flagged, and they're not put back. Discussion of how HN's flagging system works - or doesn't - is explicitly banned at the post level. You can only talk about it in comments.
So, no, that isn't just my opinion, or paranoia. And the fact that you don't know how often those stories are unfairly flagged and never put back is actually evidence of the tightness of the blinkers here. It's been crazy this past year - just try looking at my favorites.
Your favourites have tons of "If they'd cover it on TV news, it's probably off-topic" stuff. The fact that so much of it is flagged is a feature of this place, not a bug. This feature is why I keep coming here almost daily for the last 15 years or so.
Given there's a lot of the audience here being actively involved in building said Torment Nexus, it's not a surprise that discussing it can generate friction. Like Upton Sinclair so nicely put it: "It is difficult to get a man to understand something when his salary depends upon his not understanding it."
For those interested in more details and quite mind blowing examples, here is a fascinating interview with Michael Levin (one of the researchers mentioned in the article).
Start at 1:19:11, the stuff before is him talking about biology, but from an intelligence perspective. After this time stamp is his retrospective on his bioelectricity research over the years, showing also examples of how they got a frog embryo to produce eyes, and many more things.
Thanks for sharing this timestamp - Levin's retrospective on bioelectricity research is compelling. What fascinates me most is how his work challenges the gene-centric view of development. The experiments showing bioelectric patterns can override genetic instructions (like inducing eye formation in non-eye tissue) reveal a whole layer of morphogenetic information we're just beginning to understand.
> I don’t believe genetics ever claimed to provide a theory of why eyes grow where eyes grow.
That’s the whole point of developmental biology, to show how features of the human body form and develop based on gene expression, the timing of which during embryonic and fetal development itself is dictated by your genes.
If not your genes, what else would determine why you have eyes in about the same place in your head as every other human?
> The cells in your eyes have exactly the same DNA as the cells in your big toe, so developmental morphology cannot be explained with DNA alone.
Sure it can, because while every cell has essentially the same DNA, the expression of genes differs between cells, which is what causes cells to differentiate. And this differentiation also controls development; look up the Hox genes as an example.
He's changed wild-type planarians to grow the heads of other species. It reverts after a few weeks, because the system has error-correcting mechanisms, but the DNA of these worms is unchanged.
He once compared tinkering with DNA as pulling out a soldering iron to fix a software bug.
In the case of morphology, DNA may not be the best level of abstraction. It's certainly possible, just as one can use chemistry for social problems, but for some problems, affecting cell-to-cell communication may be a more direct path.
> If not your genes, what else would determine why you have eyes in about the same place in your head as every other human?
Theoretically, it could be second or much higher order effects that result from genes. It could be a combination of complex factors - the environment in the womb, nutrition, behavior by the mother, etc. - that eventually trace back to DNA.
Also, is it literally true that DNA is the only thing that's consistent (in these respects) between all generations of Homo sapiens?
Your last paragraph is their point: genes are regulated to produce that effect. The genes themselves aren’t doing it, but eg diffusion of chemical signals to inactivate genes.
Morphology is determined by the combination of genes, chemical signals, original cell machinery, and apparently electrical signals. But we never believed that genes determined morphology alone, eg, we know that chemical signals can cause anomalies.
> Morphology is determined by the combination of genes, chemical signals, original cell machinery, and apparently electrical signals. But we never believed that genes determined morphology alone, eg, we know that chemical signals can cause anomalies.
For the consistent parts - eyes may be different colors but are overwhelmingly consistent - what else could be the ultimate cause but DNA? For example, if those chemical signals, cell machinary, and electrical signals produce the same results billions of times over 200,000 years, then they must function the same overall. How does that happen if the chemical signals, cell machinary, and electrical signals aren't determined, even if indirectly, by DNA?
Your eyes would be misplaced if the process from cell clump to mat to tubule failed due to chemical signaling failure, but the whole embryo tends to be spontaneously aborted when gestation fails so catastrophically.
And despite genitalia being roughly one of two forms and similarly positioned, chemical signals can disrupt their formation.
> How does that happen if the chemical signals, cell machinary, and electrical signals aren't determined, even if indirectly, by DNA?
They don’t produce the same results with perfect accuracy — 75% of pregnancies are spontaneously aborted, at least in part due to developmental failures.
But the problem with this argument is simple: you have a human cell everywhere you have human DNA, so those correlations with DNA are also correlations with cellular machinery and with particular chemical signals from the mother. There was no point in those 200,000 years where DNA operated independently of those other mechanisms — we can only say the system as a whole reliably creates those features.
Interesting points, especially about the challenge of correlation. I guess we could remove DNA and see what happens ...
Somehow the machinary is passed down: Do we know of another mechanism besides DNA that is self-perpetuating? Is there any living creature without it? Prokaryotes (bacteria) even have DNA.
Or is there a way to do it without self-perpetuating mechanisms? Is that logically possible? Some machinary might be perpetuated by other machinary, e.g. the chemical might recreate the electrical, meaning it's not self-perpetuating. But that's not different than DNA: DNA itself isn't the machinary, but its self-perpetuation is what recreates other parts.
I suppose some parts of the environment are consistent, such as sunlight, air, water, and heat, but the environmental stimuli must trigger something that is already there.
> I guess we could remove DNA and see what happens ...
If I have a stool with three legs, and remove one leg causing it to fall, can I conclude that removed leg is what made it stand?
You’re making the same mistake as before in reverse: DNA would do nothing without a host cell or chemical signals, either.
> Somehow the machinary is passed down: Do we know of another mechanism besides DNA that is self-perpetuating?
The system as a whole is self-perpetuating, but DNA is not self-perpetuating: without a host cell and without ambient chemical signals, it cannot propagate. That’s in contrast to ribozymes which can be self-catalyzing RNA, ie, truly self-propagating chemicals.
In the RNA world hypothesis, such self-catalyzation was the origin of life; and by the time DNA evolved, it did so within a running biological system and as merely one component of cellular replication.
As a whole the system of chemical signals, DNA, and cellular machinery propagates; but just like our stool example, removing any of the factors causes that to fail.
The DNA removal comment was as joke; sorry if that wasn't clear.
No system is self-perpetuating, per the Second Law; all need other inputs. What makes the machinary yield the ~same results ~every time is DNA.
> In the RNA world hypothesis, such self-catalyzation was the origin of life; and by the time DNA evolved, it did so within a running biological system and as merely one component of cellular replication.
Is there evidence of that? Afaik the earliest evidence is prokaryotes ~~3.5 billion years ago, and prokaryotes generally have DNA.
> The cells in your eyes have exactly the same DNA as the cells in your big toe
Is that true?
I know that cells in the brain have significant variability in DNA, but not really aware of what non-neuronal and non-brain cells in general typically have.
if anyone is interest (note: i'm just a software engineer, not in biomedicine or bioelectricity yet). if there are already long running discord servers then i'd rather join these instead :)
There is a great documentary series with three archeologists by the BBC about medieval castles featuring Guédelon as a real live example from around 2014. I really enjoyed watching this and highly recommend it.
Great show. If you liked it, Ruth Goodman and Peter Ginn has made several shows together along with another archeologist, Alex Langlands, which I think are even better.
Of these, I think my favourite is Victorian Farm (2009), where the gang has to bring a real Victorian-era farm back into working order and then live like the Victorian farmers did. Unlike the castle show, it benefits from the gang having to research and learn the old ways on their own, whereas the castle is a big project where they're being taught or directed by the crew who's already working there.
The other shows — Tudor Monastery Farm, Edwardian Farm, Wartime Farm, Tales from the Green Valley, etc. — are all thoroughly excellent.
A minor point, but Goodman is not an archeologist or historian, but she's very good!
That was the show that pointed out to me that Henry VIII had much bigger problems than divorce to deal with. By his reign the Crown owned about a third of England. And so did the Catholic Church.
Now maybe the story about him breaking with the Church over their stance on divorce wasn’t completely bullshit, but he was dealing with an existential threat to both the Crown and his family line and they were both being authored by the Church.
I had some questions about the history of land ownership of the crown and the church occasioned by this comment but I guess I found answers in an article after googling (wow, it's been a long time since that has happened, feels like 2006 all over again) so I made a post on the article I found
A Short, Angry History of Land in Britain, by Thom Forester
It's actually alarming how many of the things they did to screw the peasants out of opportunity and freedom echo "landmark" legislative actions and key regulatory trends in the 20th century US.
The Tudor show pointed out that the mills tended to be owned by the church and so you pretty much had to pay them in a fraction of the take. I knew being a peasant fucking sucked but I thought it was mostly the lords to blame. Not church too.
That show was great but I think they missed an opportunity to tie the events it dramatizes into the broader geopolitical context that would greatly shape our civil wars.
In forcing "King Boats'n'hoes" to take his country and go it alone when he did and under the circumstances he did the pope kicked over one of the key dominoes in the line that leads to our modern balance of power. It's one of those pivotal moments in world history that only really happened the way it did due to the inclinations and personalities of the people involved.
I mean the tales of Robin Hood go a lot into robbing the Church, indeed Friar Tuck is first met and accused of being bad because a rich friar until he proves his mettle.
A bit offtopic. Is it possible to cast these videos somehow from the browser to a TV? I know it's possible to download them, but I am wondering if it's possible to stream them to the TV instead...
Yeah. I also wanted to rely on chromecast for this. But the chromecast button does not even show up when I open that page, so probably that page does not support that
You can still "cast tab" from Chrome. That's mostly what I use, to be honest. Some sites / streamers do a poor job implementing the cast functionality.
If you have a Chromecast on your tv, you can generally just cast a whole tab using chrome. But the frame rate will be awful and I don't think audio gets cast.
If there isn't a cast button within the video player itself, i would download them, and then use something local to cast the actual video content to your tv (vlc has this feature)
Additionally to locking the exposrue time and aperture, one could also take multiple exposures, figure out the camera's light response function and fuse multiple exposures together into a single higher dynamic range (HDR) image (see OpenCV tutorial on that or Debevec et al. 1997) Assuming you can find the camera response for the very long exposure times at night _and_ the very short during the day, one could relate them to each other and display both for accurate visual comparison.
It's a way to cut just the components/styles/themes/patterns out from a model and apply them into other models.
So if I have a Disney characters checkpoint, but I really like this MakeGiantEyes checkpoint, if I can get it down to a MakeGiantEyes LoRA, I can apply that on top of my Disney Characters model which is already a custom trained set. It definitely does not always work, but when it does it's like magic. At a practical level, it's a model-modifier.
... It took me a minute to get those because I had to sort through a LOT things that would probably get me banned here. If anyone wanted to know nationalities were using SD more... It's Asians, hands down, all day long, and I think that's interesting.
EDIT: And if you were wondering what a Textual Inversion is vs a LoRA... Don't ask me! They're both model modifiers, but as I understand it, textual inversions are good for faces (which is why most of those are people, and they are kilobytes in size), and LoRAs aren't as good for faces specifically but better for themes.
TI exmaples (I couldn't use any of the million women... there are almost none that would be appropriate to post. Even though civitai does a good job of removing the nsfw posts of real people even with cloths on some are just still too much... Thirst is driving AI now)
https://civitai.com/models/11039/ian-mckellen
or
https://civitai.com/models/8060/seu-madruga
Like, a generative model is a source of _some_ information that is refined with gates (classification models) conditional on the generated information?
The analogy to transistors and logic gates falls flat a bit when you consider that voltage is a rather simple univariate signal, while generated text is quite complex and multivariate. But I understand that the main point is the composability and filtering.
Think of it as information, not voltage. An XOR produces information. A lot of XORS with ANDs make a calculator which opens up an entire vectorspace of mathematical information.
I try to. One similar thing comes to my mind: generative adversarial networks (GANs). If I'm not mistaken this is along the line of your idea of composing single ML models to bigger information processing units.
Do you, by any chance, have links or recommendations for material to read up on architectures that do consider ML models as composable gates?
I wonder what happens when an edit can not be delivered to 1+ participant(s). Does an edit need aknowledgements from _all_ participants to be persisted? Is it just dropped?
In Zed/CRDTs, these two seem synonymous to me since the OP's article is the first I read about CRDTs that I grasped, or so I thought ;)
Doesn't the whole process assume that whenever any edit is done (insertion/deleteion/un-/redo) that the edit _eventually_ reaches all participants?
So if a single edit is believed to be delivered to all, but actually never made it to a single participant, who resends it? And if it's not resent, then the participants state is inconsistent from now on, no?
So, client that was "offline" can just ask a participant to get the edits from his own last one (aka pull), send out bis own offline edits (push) and resolve from there, I See.
You are very clear about the current limitations on data size, which I find refreshingly honest! How sensible do you find the idea to fine tune the model to a specific problem that has more than 1000 observations, by resampling the data (similar to bootstrapping) and retraining on the subsamples? As I understand it, one could fine tune the algorithm that TabPFN learned to the specific problem.
Many thanks also for open-sourcing your work and making the colab notebook, I've been playing around with that a bit.
We did try it a bit a while back, but did not have conclusive results. I expect you can bend it to perform better for larger datasets, too, but how exactly I cannot say for sure. The bootstrapping is definitely a good candidate for this.
Reading the guidelines I can't see how this is off-topic or does _not_ "[gratify] one's intellectual curiosity."
Edit: spelling
reply