Hacker Newsnew | past | comments | ask | show | jobs | submit | cm2187's commentslogin

Your friend founded OVH?

But that means you need to have a different laser pointed at every single individual aircraft right? Doesn’t really scale.

I suppose you can do time-sharing. And use mems-mirrors to quickly move the beam between different targets.


You can probably do phased arrays. (It might already be a phased array.)

Pretty sure phased array LASERs are not yet a thing.

I was not sure, but they are!

https://cga.anu.edu.au/research/activities/laser-beam-steeri...

https://www.darpa.mil/research/programs/excalibur

I guess in some ways even the fancy multi diode fiber lasers are phased arrays, just with the single goal of higher output power.


Looks like these are in early development and nowhere near ready like this test was.

Lasers are coherent emitters; you can definitely make interference patterns with them, so I don't see why LASER MIMO wouldn't be possible, in theory.

Yeah but this is research, if they're to come up somewhere, where else would it be?

If starlink satellites get laser downlink, it might work :P

laser downlink to one point, isn't it? Not to 300 moving aircrafts at once.

Accept the cookies and flush them out every time you close the browser. I think it would be naive anyway to assume that clicking no on a cookie banner would achieve much for your privacy.

So-called "cookie banners" usually ask for your consent to much more than optional tracking cookies. By accepting you might be giving your permission to e.g. track you through various fingerprinting methods, build a profile and share it with advertising partners.

If they are aggressive enough to do fingerprinting, what makes you think they would abide to your choice? You do browser fingerprinting when you want to overcome people rejecting cookies.

Usually it's the fine of 2% of total global revenue (not profit)

An additional reason for not browsing the web without uBlock Origin on Firefox or other browsers with full support (not Chrome).

Why even ask for the cookies if denying them doesn’t achieve much?

It’s naive to think that cookies are the only tool used for tracking, but they are the most powerful tool for web based tracking.


Because in some legal systems you're required to ask. You're also required to follow fairly specific rules relates to the user's selection and data, though I can't imagine enforcement keeps up with websites breaking those laws.

Because EU Cookie Law was a flawed idea?

How so? The law doesn't require cookie banners. However, you could argue that tracking/advertisement cookies should have been banned completely and that the law is flawed in that it allows for tracking given user "consent".

I love the EU apologists - “it wasn’t a bad law just because the outcome was bad”

The alternative being to bend over and grab our ankles with both hands the moment the scummy ad-tech industry requests our data?

Sorry mate, the GDPR is there for a bloody good reason; and legit companies obey the law.


The GDPR is theater. An effective privacy law would have prevented data collection in the first place. Data collected will be abused, and a cute little banner won't change this.

Ummmmm.

The GDPR does outlaw unnecessary collection of personal data without explicit opt-in consent. It's baffling you appear ignorant of this.


The consent is the problem. It should be illegal even with consent, so this whole industry wouldn't exist.

Yet, Facebook and Google have a thriving business in the EU among I’m assuming EU companies.

So you can pretend that the law is effective or you can admit that all it gave the world were cookie banners.


The "cookie banners" allow you to opt out of providing your personal data. That is the entire point!

Blame the parasitic ad-tech industry for their existence. Not the lawmakers who protect all our privacy.


Yes because of the GDPR, there aren’t still two trillion dollar+ market cap ad Tech companies.

But at least we have cookie banners everywhere.


More pity to those who (for some bizarre reason) voluntarily choose to interact with those ad-tech companies.

So you don’t use Google and don’t have an Android phone?

It was not a flawed idea, but flawed execution. The law should have mandated to adhere to the user's "do not track" setting in the browser.

That being said, it was very early regulation in this field, and more recent approaches are already better, e.g., GDPR, DMA.


No, shan’t give them the metrics :)

Retire under a bridge...

Did they stutter?!

so my smart microwave will require some age verification?

Of course! Think of the dangers of an unsupervised child... (SHOCK WARNING) cooking... A gasp MEAL!

The original notepad was completely broken in the first place anyway. Can't handle large files, ctr+Z would cancel a large random number of previous actions, the search feature is case sensitive and barely useful, etc. What do you miss vs a Notepad++?

I absolutely need a pure SDI workflow. I like browsing through files on disk by the double-clicking them, reading through them a bit, and then using Alt+F4. I tried that with Notepad++ once and I ended up with hundreds of files open in tabs because Notepad++ remembers everything you've ever opened even if you Alt+F4 the app.

I also like sometimes having multiple files open at once and drag the windows around my monitors as I need to and you can't do that in Notepad++.

Also Notepad++'s UI is bloated. There's just too much going on.

And please don't suggest I muck around in settings, I'm not interested in spending hours mucking around in a gigantic bloated settings dialog, I want something that Just Works™ with no configuration just like Win95 Notepad.

Edit: My specific use case here is for viewing files, not editing them. I use a different editor when I'm actually writing stuff but for browsing I just want old Notepad.


What is the version used by the free chatgpt now? (https://chatgpt.com/)

> Since the car wash is only 50 meters away (about 55 yards), you should walk.

> Here’s why:

> - It’ll take less than a minute.

> - No fuel wasted.

> - Better for the environment.

> - You avoid the irony of driving your dirty car 50 meters just to wash it.

the last bullet point is amusing, it understands you intend to wash the car you drive but still suggests not bringing it.


> You avoid the irony of driving your dirty car 50 meters just to wash it.

The LLM has very much mixed its signals -- there's nothing at all ironic about that. There are cases where it's ironic to drive a car 50 meters just to do X but that definitely isn't one of them. I asked Claude for examples; it struggled with it but eventually came up with "The irony of driving your car 50 meters just to attend a 'walkable neighborhoods' advocacy meeting."


That's actually an amusing example from Claude.


> it understands you intend to wash the car you drive but still suggests not bringing it.

Doesn't it actually show it doesn't understand anything? It doesn't understand what a car is. It doesn't understand what a car wash is. Fundamentally, it's just parsing text cleverly.


By default for this kind of short question it will probably just route to mini, or at least zero thinking. For free users they'll have tuned their "routing" so that it only adds thinking for a very small % of queries, to save money. If any at all.


I don't understand this approach. How are you going to convince customers-to-be by demoing an inferior product?


Because they have too many free users that will always remain on the free plan, as they are the "default" LLM for people who don't care much, and that is a enormous cost. Also the capabilities of their paid tiers are well known to enough people that they can rely on word of mouth and don't need to demo to customers-to-be


They're not more default than people innocently googling something and getting an AI response from some form of Gemini.


Right, but that form of Gemini is also not the top Gemini model with high thinking budget that you would get to use with a subscription, the response is probably generate with Gemini Flash and low thinking.


Through hype. I am really into this new LLM stuff but the companies around this tech suck. Their current strategy is essentially media blitz, reminds me of the advertising of coca cola rather than a Apple IIe.


It's all trade offs. The router works most of the time so most free users get the expensive model when necessary.

They lost x% of customers and cut costs by y%. I bet y is lots bigger than x.


The good news for them is that all their competitors have the exact same issue, and it's unsolvable.

And to an extent holds for lots of SaaS products, even non-AI.


I don't understand why they need to save money...


Every business needs to minimize costs in order to maximize profits.


I think this shows that LLMs do NOT 'understand' anything.


> I think this shows that LLMs do NOT 'understand' anything.

It shows these LLMs don't understand what's necessary for washing your car. But I don't see how that generalizes to "LLMs do NOT 'understand' anything".

What's your reasoning, there? Why does this show that LLMs don't understand anything at all?


I think this rather shows that GPT 5.2 Instant, which is the version that he most probably used as a free user, is shit and unsusable for anything.


Another/newer/less restricted LLM may give a better answer but I don't think we can conclude that it 'understands' anything still.


If it answers this out-of-distribution question correctly -- which the other major models do -- what else should we conclude, other than that a meaningful form of "understanding" is being exhibited?

Do we need a new dictionary word that acts as a synonym for "understanding" specifically for non-human actors? I don't see why, personally, but I guess a case could be made.


You may be tempted to conclude that. Then you find something else to ask that leads to an answer obviously nonsensical to a human being, or it hallucinates something, and you realise that, in fact, that's not the case.

IMHO 'understanding' in the usual human sense requires thinking and however good and fast improving LLMs are I don't think anyone would suggest that any of them has become sentient yet. They can infer things based on their training data set better and better but do not 'understand' anmything.

This is a deep and complex topic, and has been for decades.


Gemini 3 Flash answers tongue-in-cheek with a table of pro & cons where one of the cons of walking is that you are at the car wash but your car is still at your home and recommends to drive it if I don't have an "extremely long brush" or don't want to push it to the car wash. Kinda funny.


As long as there is a forum as technical as this where LLM performance commentary uses the word "it understands" irony is still alive.

There is no problem that cannot be solved with creating a bureaucracy and paperwork!


I understand this is tongue-in-cheek, but do you have an alternative/better proposal?


Let the market do. If good data is so critical to the success of AI, AI companies will pay for it. I don't know how someone can still entertain the idea that a bureaucrat, or worse, a politician, is remotely competent at designing an efficient economy.


All the world's data was critical to the success of AI. They stole it and fought the system to pay nothing. Then settled it for peanuts because the original creators are weak to negotiate. It already happened.


No they won't pay for it, unless they believe it's in their best interests. If they believe they can free-ride and get good data without having to pay for it, why would they lay down a dollar?


Because the companies in control of that data won't let them have it for free, like what is happening in the article.


Or, they'll just create more technically sophisticated workarounds to get what they want while avoiding a bad precedent that might cost them more money in the long run. Millions for defense, not one cent for tribute.


Now apply the same logic to laws, except that laws are a lot slower to change when they find the next workaround.

And it's a lot harder to get the law to stop doing something once it proves to cause significant collateral damage, or just cumulative incremental collateral damage while having negligible effectiveness.


For the academy awards, to its defense, it was competing against Pulp Fiction, Forrest Gump, Four Weddings and a Funeral, or the Madness of King George. I can barely name one good movie a year these days, and certainly none that makes it to the oscars. The contrast with the 90s is brutal.


> can barely name one good movie a year these days

Not really.

Of the recent movies, Everything Everywhere All at Once is a storytelling masterpiece. Since you mentioned it, I personally rate it alongside Tarantino's Pulp Fiction.


Everything Everywhere All at Once was the last time I sat in a theater where, for the first half at least, I thought I was watching an instant classic.

But that movie just dragged on, and now I look back and see it as a bungled opportunity. It could've been so much tighter in the edit. They could've cut a third of the movie and made the whole thing so much better.


This has generally been my experience with most highly acclaimed movies over the past 10 years. Most recently had this w/ Marty Supreme... last year had this w/ The Brutalist and The Substance.

The first half has me thinking instant classic, my hope is sky high. But then toward the end I find myself looking at my watch and realize it's simply not going to the stick the landing.

OTOH, many acclaimed streaming series have generally done this well. My take is that as long-form storytelling has evolved, movies have transitioned into this post-modernist phase as directors/writers don't feel they have the runway to tell something truly cohesive that doesn't end up being trite. It's much more about saying 'something' or imbuing a feeling than telling a fully fleshed 3 act story.


I did like how The Brutalist at least included an intermission the way long movies used to do.


> They could've cut a third of the movie and made the whole thing so much better.

I feel that way with Inception. That out of nowhere 30-minute snow action part dragged on forever.


> They could've cut a third of the movie and made the whole thing so much better.

It would become just an action movie with crazy plot then.


> Everything Everywhere All at Once is a storytelling masterpiece

I thought it was so awful I gave up half way through. Maybe it gets better after that. But I agree on Pulp Fiction.


Same here. I got extremely annoyed with the constant ridiculous fight scenes. About halfway through I gave up.


It's Children of Men crossed with The Big Lebowski, with Pynchon instead of noir characters. When you get what it's trying to do, it gets better.

I bounced off of it at first, but I bounced (hard) off of Lebowksi as well.

I don't think it's PTA's best film (or that I will come around to that opinion eventually), but it's pretty good.


PTA did not make Everything Everywhere All at Once.


Oof, I got my wires crossed. Nothing I said makes any sense in the context of EEAaO. I was thinking of OBAA. Thanks!


Me too. Extremely loud, lots of flashing and fast cuts.

I genuinely didn’t really think there was a story, just spectacle.


Agreed. Not as good a film as it was advertised to be.


Me too. And love pulp fiction. Just used Mr. Wolf to reference a situation at work.


The "OK, let's not start sucking... yet " is the one that comes to mind during production fires but can't use that one at work unfortunately


This was a good movie, but what was it up against. Were there 4 or 5 other movies of comparable goodness that any of could have won the oscar? So 'can barely name one good movie' is apt here. There are some, but way fewer and farther between.


Everything Everywhere... is a much better movie than the incredible Pulp Fiction. Some of the visual effects are actually psychedelic (I've "seent" them), and the storytelling is exceptional.

The scene where the antagonist is walking down a hallway while the background keeps changing — is among the best fight scenes / visuals in any film, ever.


I think you're going to see more and more people saying things like that as the audience gets younger and more people see the antecedents of Pulp Fiction before they see Pulp Fiction itself. There wouldn't be an EEAaO without Pulp Fiction.

Even setting its influence aside, Pulp Fiction is the better movie.


I wouldn't even rate pulp fiction highly on Tarantino's filmography. I tried watching it recently and found it to be incredibly pretentious and overwritten.


It's quintessential-Tarantino, but I don't ever recommend it anymore (start with Django or Reservoir Dogs). Decades ago I shared this movie with college friends — mostly because we enjoyed decadence.

If you've not seen Pulp Fiction by 2026 [0], how can I safely recommend you submit yourself to hours of semi-disconnected robberies, rapes, and deceit? It's a great movie, EEAaO is just better storytelling.

[0] similarly, how does one recommend the acclaimed Deliverance without blushing?


Funnily enough I deeply dislike Django too and think Reservoir Dogs is good but extremely raw and unrefined.

My favourites of his are probably Once upon a time, Jackie Brown and Deathproof, with an honourable mention for Basterds.


=D

Django has low re-watchability (unlike most of Tarantino's work) but incredible acting/twists/cinematography.

Once Upon a Time is too much for me (bottom-tier Tarantino IMHO), but it does have many great actors/scenes (the overall storyline/premise is what I didn't care for).

Haven't seen Deathproof, but Basterds is wonderful storytelling.


Yeah I think Basterds is probably the most undeniably great, even if it's not my favourite. He even calls his shot with the last spoken line being “i think this might be my masterpiece”.

Probably my favourite thing about cinema is how slippery the subjective experience is.

For example I can appreciate a movie I don't really enjoy in a way I can't with music. Also on a rewatch a movie can go from hated to loved, or vice versa, in a way that feels unique to the medium.


>Yeah I think Basterds is probably the most undeniably great [Tarantino film], even if it's not my favourite.

Well-said.

>...on a rewatch [it] can go from hated to loved

I typically don't rewatch movies for at least five years — this is enough time for life experiences to change media interpretations. Yet I listen to the same tracklist of catchy MP3 earworms, on repeat.

Songs are motivational background energy (for me), and skipping a track isn't nearly as hard as bailing out of two hours invested in a cozy full-length film.


> There wouldn't be an EEAaO without Pulp Fiction

How so? This is an intriguing statement and I want to hear more.


>>more people see the antecedents of Pulp Fiction before they see Pulp Fiction itself.

Can he also explain this above statement, please?



Here are the only notables I can think of since Everything Everywhere

Triangle of Sadness https://www.imdb.com/title/tt7322224

Coming Home in the Dark https://www.imdb.com/title/tt6874762


You can only think of 2 notable films since 2022?

* American Fiction

* The Holdovers

* Oppenheimer

* Perfect Days(!)

* Nosferatu

* Conclave

* Challengers

* The Mastermind

I can rattle off more but those seem pretty hard to argue with. All of them are better than EEAAO.


I just watched "Frankenstein" (by Guillermo del Toro) from last year, and thought it was pretty fantastic.

"Flow" in 2024 was also fantastic.


Just wanted to second "Flow" - undoubtedly one of the best animated movies of all time! Give it a try if you can, you won't regret it.


Make sure to make a place for your cat on the couch too: he or she will probably love it.


I found Nosferatu to be a snooze fest (watched it in the theater by myself so I could take it all in) but maybe it needs a rewatch.


I thought it did an extremely good job of conjuring a particular place/time, and I find the Nosferatu backstory of being Temu Dracula sort of inherently entertaining.


Goodness no. It was such a drag! That movie became famous from the hype. I couldn’t finish it. I am really wary of famous + acclaimed films now. These days this combo almost always disappoints. Like Nolan films. I know he has a massive “fan base” now and anything he churns out will become crazy famous and an instant classic. Anything!


Last year's winner Anora was also excellent.


YMMV. I found Anora quite tiresome - all of the people depicted were awful and stupid, and the point that it made was so basic that it could have been made in 10 minutes flat. I'd call it "preachy" but that's overselling it.


Fair enough, not everyone needs to like the same things. In fact, I had a rather negative view on Shawshank Redemption, but it's been too long since I saw it that I barely remember why.


> > can barely name one good movie a year these days

> Not really.

Not really meaning you can't really name one good movie a year (i.e., agreeing with OP)? Because your example of a good recent movie was 4 years ago.


EEAO was four years ago now.


That’s only one movie.


Funny, I thought it was absolutely terrible.


YMMV. I found EEAAO to be engaging but shambolic. It was an experiment that kinda worked, kinda not. The chaos of it can't be cleaned up, it's intrinsic to the concept.

It's not going to a template for lots of similar films. It's more of a one-off.

But anyway, that was several years ago, it stretches the meaning of "recent".


Train Dreams


Unbelievable film. I am so appreciative this was made.


Well you either remove all the bureaucracy around drug testing and approval and make it cheap to develop a new drug, or you prevent drug makers from making money if they are successful at developing a new molecule. But if you do both, all you will get is zero research. Right now it takes 10s of billions in R&D budget to bring new molecules to the market, which is insane.


Or you pay them directly. Most countries have research funding. Since there's no way to know what you'll find or how long it will take, research doesn't fit well in the capitalist model. Makes much more sense to apply a fixed effort and accept whatever results come out, but only the government can do that — or a rich monopoly like Bell.


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

Search: