I disliked them a bit, but then they stopped flying to a certain destination. I quickly realized that the other airlines were 3x more expensive. I realized I actually cared about price much more than any possible extra leg room or other perks, and that their super cheap flights are quality by itself.
Because LLMs are based on the abstract ideas of neural nets from brains. Say what you wish, but some problems were completely unsolvable before we adopted this paradigm. On some level, we must've gotten some ideas close to the right ballpark.
I had the same confusion initially, interestingly chat GPT gets it:
So while wolfgang42 wasn't there when Ulbricht was actually arrested, their realization created a vivid mental image of the event unfolding in that space, which made the story feel more immersive.
In short: they were reading about an old event, but it happened to occur in the same spot they were sitting at that moment. Hope that clears it up!
> their realization created a vivid mental image of the event unfolding in that space, which made the story feel more immersive.
Glad that ChatGPT, probably like GP themselves, is a visualizer and actually can create a "vivid mental image" of something. For those of us with aphantasia, that is not a thing. Myself, I too was mighty confused by the text, which read literally like a time travel story, and was only missing a cat and tomorrow's newspaper.
Legitimately and I say this was absolutely no shade intended. This is a reading comprehension problem, nothing to do with aphantasia.
He clearly states that he was reading an article, he uses past tense verbs when referring to Ross, and to the events spelled out in the article. If you somehow thought that he could be reading an article that ostensibly has to be describing a past event as he was seeing it in real time that is a logic flaw on you.
It has nothing to do with what you can or cannot visualize. All you have to do is ask yourself could he have been reading an article about Ross’s arrest while watching it? Since nobody can violate the causality of space time the answer is no.
This isn’t just you this is everybody in this thread who is reading this and going this is a little confusing. No it’s very clearly him speaking about a past experience reading an article about a past event.
I realised what was going on, but I did a double-take at:
> Then Ulbricht walked into the public library and sat down at the table directly in front of me
The problem is that two past events are being described, so tense alone cannot distinguish them. Cut the readers some slack; the writing could have been better.
Done for effect: it felt to the OP as if it was the present so the writing conveys that, while elsewhere making it clear the arrest was not the present.
I am as baffled at the responses and appreciated this explanation as it was helpful to me to work on my communication style and expresses a lot of similar frustrations I have. Like what is actually going on here? this isn’t shade at anyone, I just feel like people are losing some fundamental ability to deduce from context what they are reading. it’s doubly concerning because people immediately reach to an AI/LLM to explain it for them, which cannot possibly be helping the first problem.
Agree. This entire thread is weird. How do so many people in this thread have such obvious reading comprehension issues?
On a similar note--I've noticed that HN comments are often overwrought, like the commenter is trying to sound smarter than they actually are but just end up muddling what they're trying to say.
I wonder what is going on? I’ve noticed this getting worse for a long time to the point I’m not sure it’s my imagination anymore. I usually like to lambast whole word reading as a complete failure in the american school system that contributes to this, but I think it’s likely something else. Shorter attention spans?
We have a multitude of immediate distractions now.
Books build richer worlds & ideas. But without learning to love books very early in life, which takes a lot of uninterrupted time, they don’t come naturally to most.
I used to read a few books a week, virtually every week. Sometimes two or three in a long day and some night. I still read a lot daily, interesting and useful things in short form. But finding time to read books seems to have become more difficult.
I do think the comment had something about how it was written that made it hard to follow. I understood the first sentence. But then I got to
> Having this tableau unexpectedly unfold right in front of my eyes
And the metaphor / tense shift caught me by surprise and made my eyes retrace to the beginning. I still got it, but there was a little bit of comprehension whiplash as I hit that bump in the road.
In some ways, we're treated to an experience like the author's as we hit that sentence, so in that sense it's clever writing. On the other hand, maybe too clever for a casual web forum instead of, say, a letter.
Right. I'm not claiming the LLM has visual imagination - I suspect that OP has it, and that ChatGPT was trained on enough text from visual thinkers implicitly conveying their experience of the world, that it's now able to correctly interpret writing like that of OP's.
It's a strange feeling, watching the AI get better at language comprehension than me.
I made a similar mistake on the original comment as you (I read it as "Ulbricht returned to the cafe, he actually sat down right in front of me while I was reading the story about his previous arrest here, and that's when I realised it was the same place"), and also thought you were saying that you think ChatGPT has a visual "imagination" inside.
(I don't know if it does or doesn't, but given the "o" in "4o" is supposed to make it multi-modal, my default assumption is that 4o can visualise things… but then, that's also my default assumption about humans, and you being aphantasic shows this is not necessarily so).
You could also say that ChatGPT erred similarly to the original writer, who was unclear and misleading about events.
We needn't act like they share some grand enlightenment. It's just not well expressed. ChatGPT's output is also frequently not well expressed and not well thought out.
There's many more ways to err than to get something right. ChatGPT getting OP right where many people here didn't tells us it's more likely that there is a particular style of writing/thinking that is not obvious to everyone, but ChatGPT can identify and understand, rather than just both OP and ChatGPT accidentally making exactly the same error.
Why would that be more likely? Seems like OP and ChatGPT (which is just many people of different skill levels) might easily make the same failure to communicate. Many failures of ChatGPT are failures to communicate or to convey structured thinking.
Because out of all possible communication failures OP and ChatGPT could make, them both making the exact same error, in a way that makes the two errors cancel out, is extremely unlikely.
Reducing any judgment out of your comment, you have to admit that the commenter's action was a successful comprehension strategy they learned from and can use in the future without chatgpt.
Okay, that's actually pretty wild. I totally misunderstood too, but the response from the "AI" does indeed "clear it up" for me. A bit surprised actually, but then again, I suppose I shouldn't be, since language is what those "large language models" are all about after all... :)
Indeed. But their is something surprising here, however. people like chomsky would present examples like this for decades as untracktable by any algorithm, and as a proof that language is a uniquely human thing. they went as far as to claim that humans have a special language organ, somewhere in their brain perhaps. turns out, a formula exists, it is just very very large.
> chomsky would present examples like this for decades as untracktable by any algorithm, and as a proof that language is a uniquely human thing
Generatove AI has all but solved the Frame Problem.
Those expressions where intractable bc of the impossibility to represent in logic all the background knowledge that is required to understand the context.
It turns out, it is possible to represent all that knowledge in compressed form, with statistical summarisation applied to humongous amounts of data and processing power, unimaginable back then; this puts the knowledge in reach of the algorithm processing the sentence, which is thus capable of understanding the context.
Which should be expected, because since human brain is finite, it follows that it's either possible to do it, or the brain is some magic piece of divine substrate to which laws of physics do not apply.
The problem turned out to be that some people got so fixated on formal logic they apparently couldn't spot that their own mind does not do any kind of symbolic reasoning unless forced to by lots of training and willpower.
That’s not what it means at all. You threw a monkey in your own wrench.
The brain has infinite potentials, however only finite resolves. So you can only play a finite number of moves in a game of infinite infinities.
Individual minds have varying mental technology, our mental technologies change and adapt to challenges (not always in real time.) thus these infinite configurations create new potentials that previously didn’t exist in the realm of potential without some serious mental vectoring.
Get it? You were just so sure of yourself you canceled your own infinite potentials!
Remember, it’s only finite after it happens. Until then it’s potential.
No, it doesn't. The brain has a finite number of possible states to be in. It's an absurdly large amount of states, but it is finite. And, out of those absurd but finite number of possible states, only a tiny fraction correspond to possible states potentially reachable by a functioning brain. The rest of them are noise.
You are wrong! Confidently wrong at that. Distribution of potential, not number of available states. Brain capacity and capability is scalar and can retune itself at the most fundamental levels.
As far as we know, universe is discrete at the very bottom, continuity is illusory, so that's still finite.
Not to mention, it's highly unlikely anything at that low a level matters to the functioning of a brain - at a functional level, physical states have to be quantized hard to ensure reliability and resistance against environmental noise.
Potential is resolving into state in the moment of now!
Be grateful, not scornful that it all collapses into state (don’t we all like consistency?), that is not however what it “is”. It “is” potential continuously resolving. The masterwork that is the mind is a hyoerdimensional and extradimentional supercomputer (that gets us by yet goes mostly squandered). Our minds and peripherals can manipulate, break down, and remake existential reality in the likeness of our own images. You seem to complain your own image is soiled by your other inputs or predispositions.
Sure, it’s a lot of work yet that’s what this whole universe thing runs on. Potential. State is what it collapses into in the moment of “now”.
And you’re right, continuity is an illusion. Oops.
I never bought into Searle's argument with the Chinese room.
The rules for translation are themselves the result of intelligence; when the thought experiment is made real (I've seen an example on TV once), these rules are written down by humans, using human intelligence.
A machine which itself generates these rules from observation has at least the intelligence* that humans applied specifically in the creation of documents expressing the same rules.
That a human can mechanically follow those same rules without understanding them, says as much and as little as the fact that the DNA sequences within the neurones in our brains are not themselves directly conscious of higher level concepts such as "why is it so hard to type 'why' rather than 'wju' today?" despite being the foundation of the intelligence process of natural selection and evolution.
* well, the capability — I'm open to the argument that AI are thick due to the need for so many more examples than humans need, and are simply making up for it by being very very fast and squeezing the equivalent of several million years of experiences for a human into a month of wall-clock time.
Minds shuffle information. Including about themselves.
Paper with information being shuffled by rules exhibiting intelligence and awareness of “self” is just ridiculously inefficient. Not inherently less capable.
I don’t think I understand this entirely. The point of the thought experiment is to assume the possibility of the room and consider the consequences. How it might be achievable in practice doesn’t alter this
The room is possible because there's someone inside with a big list of rules of what Chinese characters to reply with. This represents the huge amount of data processing and statistical power. When the thought expt was created, you could argue that the room was impossible, so the experiment was meaningless. But that's no longer the case.
I'm not sure I'm following you. My comment re Chinese room was that parent said the data processing we now have was unimaginable back in the day. In fact, it was imaginable - the Chinese room imagined it.
Well, simple there is no need to reconcile it ? One can be successfull in one area and fail in another.
Tesla is not a bad company. As an EV maker, it's as good as it gets (although it's still bad as a business). But if it was just that, it would be valued at 10% of what it is today, at best.
But it's promoted by Musk, and priced by the market, as a revolutionnary one. And if you look at the history of it, you see it's all smoke and mirrors. In the process Musk cashed out > 40 billion USD.
They don’t seem to be a great EV maker anymore… they were definitely first to market with practical luxury EVs, but the old school car companies have caught up with better designed EVs and much higher build quality at the same price points. The high end German EVs from Porsche, Audi, BMW and Mercedes are overall much better cars then a Model S in the same price range. Not just the general build quality, but the EV tech itself is generally more advanced than what Tesla is offering- faster charging, better battery cooling and charge management, better traction control, etc.
The range of the Porsches still sucks, go to any dealer and notice the beautiful and fancy Porsche Taycan EVs sitting around unclaimed with a sales guy eager to make a deal. The BMW's are also a blight upon the eye, it's laughable.
But you aren't wrong, Kia has nailed it where it matters.. their latest are extremely popular, there are tons in the Bay Area. Yuck, but it's true*. May be worth a shot if you're in the unfortunate position of needing new transportation.
* I have no vested interest in any car company, I drive very few miles per year and my cars are old AF. I have a lot of friends, though. Only want the best for you, Internet friend. Cheers.
P.s. Combustion Kia's suck ass, trash trying to move upmarket, lol. Hello humanity.
My daily driver is a 25 year old Porsche sports car and it has less fuel range than a new Taycan. Porsche has always sold track ready cars from the factory- with all of the compromises that entails. The Taycan is a much better designed and made car than a Model S. Porsche also is a company that doesn’t exaggerate their specs like Tesla does- a quick google search suggests the real world range of the Taycan is actually higher than the Model S despite a massive difference in the specs.
SUVs and crossers are only Porsches in name/badge. If it isn't a sedan contraption it's really a VW. Good styling, but not true to the essence.
I am in admirer, but my entire collection of multiple automobiles is worth less than 25 grand. Yet they're actually pretty nice and cool, one is comfortable.
As a longtime Porsche fan that was exactly my first thought also when Porsche started making SUVs, but I couldn't have been more wrong. They're not just very much in the spirit of Porsches history- I find the SUVs to be more impressive and interesting than any of their sports cars.
The Porsche SUVs are really unique and impressive performance vehicles in their own right, each in different ways- especially the Cayenne. The Cayenne is indeed very similar to a VW Touareg, but the Touareg itself is nothing like other VWs and started out as a crazy expensive Piech era offroad supercar with V12 and V10 engines that is also the basis for the modern Bentley and Lamborghini SUV supercars. Look at YouTube videos you will see these same vehicles posting sports car worthy lap times on a track, as well as easily doing some of the most technical offroad trails. To me it's a mind blowing feat of engineering to make a single vehicle that can handle such opposed things, and do them both extremely well.
You can also find a lot of nice older Porsches for much less than 25 grand- 10 grand will get you a nearly perfect condition low mileage first gen Boxster or Cayenne these days.
So I can buy a VW that is identical to a Porsche with the same handling, build quality, interior quality, performance, charging etc for a fraction of the price.
Please tell us which model it is as this could shake up the industry.
Touareg- which VW pulled from the US market because it also cost as much, or sometimes even more than the very similar Cayenne, and people wouldn't pay Porsche money for a VW badge. In 2004 you could buy a new VW twin turbo V10 Touareg for 60-70k, or about 100-120k adjusted for inflation. VW went through a weird time when ran by Ferdinand Piëch, the grandson of Ferdinand Porsche who was involved with a lot of really high end and unusual performance cars that were cool but didn't really end up having a market.
The Macan does share a similar design to the Audi Q5, but is also heavily modified for performance over the Q5, and it isn't similar to any of VWs models. That also seems like hardly an insult as the Q5 is also a high end SUV with great performance and handling.
The history is important because Porsche and VW have been collaborating on cars since they were founded by the same person- and the “that’s not a real Porsche, just a VW” snobbery has been applied to almost every Porsche model since, even the now iconic 912. VW also occasionally makes cars that are at least as high end as anything Porsche makes.
Both of the Porsche SUVs I mentioned are sold with all 3 drivetrains BTW: electric, plug in hybrid, and pure EV.
The software on the high end euro in particular is light years ahead- and I don't mean the touch screen UIs (which I personally think are dangerous and should be illegal as you have to take your eyes off the road to operate), but the actual software controlling the cars hardware, like the traction control and stability systems, which work much better on those cars, especially VW/Audi/Porsche.
On paper Tesla dominates in range and performance specs, but they seem to be lying, while the older brands are under-estimating, as real world range and track lap time specs don't reflect the supposed higher specs of Teslas cars. Tesla seems to be a fundamentally dishonest company, whose claimed technical advantages are all lies/hype.
For example, Mercedes advertises their EQS at 350 miles range, but car magazine tests have gotten 400-500 miles on them[1]. Tesla claims 405 miles range on the Model S long range, but car magazines only get ~320.
> actual software controlling the cars hardware, like the traction control and stability systems, which work much better on those cars, especially VW/Audi/Porsche
im gonna need a source on that lol how do you even quantify that
In the last year, I've driven almost every new EV available for under $100,000.
Tesla has more range per dollar than almost any others. Many have antiquated, unusable infotainment systems designed for legacy cars. Some get marked up by $10,000 or more by dealers. None have the simplicity of charging.
Tesla is still selling most of the EVs on the road. Slowdowns attributed to Musk's shit behavior are never compared with their peers, who are having even worse slowdowns. Having driven nearly all of them, I understand why.
I bought a P3D in 2018 and it was quick. But the interior deteriorated in like six months. Rattled like crazy, center console lid wouldn’t stay closed, and after the hw3 upgrade it rattled even more.
When I made a service appointment through the app I got a message that it would cost me ($180 I think?) to have it inspected to determine where the noise was coming from.
I got FSD for $2k back then and while I only had the car for 3 years and drove mostly the 101, I found it helpful in traffic jams. However I didn’t trust it much at higher speeds because it would phantom brake or swerve randomly especially around bridges or when I was in lanes that intersected. Really scary!
I got a Mazda CX-30 after selling the Tesla and it had no interior rattles and felt more premium even though it was $26k. The steering wheel felt more like a sports car than the P3D.
Just my opinion.. I do miss the crazy acceleration. I don’t miss the rattles and poor interior quality.
Not sure if I’ll get another EV until Mazda or Honda make a nice reliable car. I’m a little scared to buy any car from companies that are burning money like crazy lol. I don’t mind moving fast and breaking things on the browser, but not on the road!!
I felt the same. In the last year I test drove the Model 3 (pre-highland), Y, and pretty much all their competitors. I still landed on a Y, and I love it. I'm always surprised at how much hate Tesla gets ... I understand some folks won't prefer their minimalist interiors, and that's fine.
>Tesla has more range per dollar than almost any others.
They don't - they inflate their estimated mileage across the board. Lucid runs circles around them dollar for dollar.
>Many have antiquated, unusable infotainment systems designed for legacy cars.
They pretty much all have carplay and android auto which runs circles around Tesla's antiquated interface. It was revolutionary a decade ago, it now looks like a tired android skin. Have you tried youtube music? You can't even properly like or dislike a song, much less properly navigate playlists.
>Some get marked up by $10,000 or more by dealers.
This literally isn't a thing. I don't know of ANY EV that you can't get at cost if you spend even 30 seconds looking. The dealers marking up cars ridiculous amounts during COVID hasn't been a normal thing for quite some time.
>None have the simplicity of charging.
Huh? You can drive a Ford Mach-E up to a Tesla charger, attach your adapter, plug it in, and it starts charging. The very first time you ever charge a given car, you do need to start it from the app to tie a tesla account to the car itself, after that it's plug and charge.
>Tesla is still selling most of the EVs on the road. Slowdowns attributed to Musk's shit behavior are never compared with their peers, who are having even worse slowdowns. Having driven nearly all of them, I understand why.
Having driven nearly all of them, I think I'm questioning if you have. Literally every other car I have driven has a nicer interior, a better technology system, and basic functionality that actually works like: auto wipers, cruise control that doesn't phantom brake for no reason, things you just take for granted in any modern car that Tesla decided wasn't worth spending a couple dollars per car to have work reliably.
Tesla is absolutely compared to peers, and no their peers aren't having "even worse slowdowns".
We realize you have a lot of bias but I think what the OP was getting at is that Tesla still has the best EV experience. The drivetrain is generally rocksolid, charging is a solved problem, going on road trips the navigation works seamlessly with optimal charging stops. The interiors are absolutely bare bones and will rattle, though I have heard the latest refresh on the 3 is a lot nicer. You are also right, they have a lot of hubris, the wipers are a debacle, it would be nice if they just paid for the sensors like all other car companies for automatic wipers.
I think the Hyundai is getting close and the Mercedes lineup is nice but a little $$$. Tesla still is a nice sweet spot.
Sure. Let’s discuss why a 6-figure Tesla doesn’t have functioning automatic wipers? Elon insisted that would be easy with nothing but cameras. And if they can’t even get the wipers working, why should any of us believe they’ll solve full autonomy with nothing but cameras?
Cybertruck became the 3rd best selling EV in the US after Model Y and 3. Looks like the people have spoken after looking at "literally almost any other EV in the last few years''.
But it has the worst build quality, interior and controls. These are not numerically measurable but they certainly do matter to prospective buyers. As does even more subjective things like styling, which is now pretty dated.
Either way I suspect it now doesn't really matter because this is becoming the "Trump-mobile" which makes it a non starter for a lot of politically conscious peoples.
counterpoint; love the interior. all the widgets in other OEMs look hilarious by comparison. controls? the car literally drives me around town by itself
Space is hard. If you think SpaceX exaggerated their timeline, then NASA (JWST, SLS, Artemis), Blue Origin, ULA, Rocket Lab, Arianne all did. See how late their rockets have been?
He's just the rich clown atop of the companies putting pressure and maybe unreasonable requests to smart people under him that as you go down the chain are actually closer to being worth their net worth.
>How do you reconcile his "bullshit" with the fact his spaceships are saving astronauts from space?
this reminds me of a common political tactic. "Mr. SoandSo couldn't be lying, look at his donation records and philanthropy! Look at the jobs he produced! The profit he has generated!"
Well, the matter-of-fact is that there is no need to reconcile the two states. He's a highly effective individual who has had great success in certain fields -- he's also a huckster con-man that over-promises and under-sells.
One could argue that his effectiveness on one sided is aided by his willing to take ethical shortcuts with his portrayed opinions and 'honesty' on the other side.
There are no pure shades of color in the human psyche.
I think fans of his would just call it '4D chess' and refer to his genius. Whatever floats your boat.
Taking your idea into straight non-sense : how many days does this space mission add to our expectation on FSD delivery? Are there grounds for a class-action lawsuit against Boeing originating from Tesla owners that are unhappy with the additional stress to the time-line? Sounds ridiculous from the other side...
>How do you reconcile his "bullshit" with the fact his spaceships are saving astronauts from space?
Easy - they aren't "his spaceships". He funded a bunch of other really smart people doing incredible things with rockets. That doesn't really exonerate him from continually spewing bullshit. If you want to go with Space-X bullshit - in 2016 he said there would be manned missions to Mars in 2024. He did it again in 2017, and multiple times after that. If there's one thing you can be sure of, whatever date Elon tells you something is going to happen, you should promptly throw into the trash where it belongs and try to find someone else at the company in question that isn't a pathological liar to give you a realistic timeline.
Wow! I love this take. Somehow with all this evidence of COT helping out LLMs, I never thought about using it more myself. Sure, we kind of do it already but definitely not to the degree of LLMs, at least not usually. Maybe that's why writing is so often admired as a way to do great thinking - it enables longer chains of thoughts with less effort.
I think it's because OpenAI's leadership lacks good taste and talent. Realistically, they haven't shifted the needle with anything really interesting in 2 years now. They're using the inertia well but that's about it. Their model is not the best, the UI is not the best, and their pace of improvement is not great either.
I find the chatgpt-4o advanced mode to absolutely be "really interesting". And the video input they showed in the demos (and hope would same day release) could be a real game changer. One thing I would like to try, once that's out, is to put a computer with it amongst a group of students listening to a short lecture about something outside its training set and then check how the AI does on a comprehension quiz following the lecture - my feeling is that it would do significantly better than the average human student on most subjects.
Except last I heard it's getting merged with the hated React router and it's not really clear what that implies, no? Haven't used either, just reading.
