Hey - plugged this into chatGPT 5.2 and it seems to think this theory needs more work.
“As written, this looks closer to sophisticated curve-fitting (numerology with constraints) than a legitimate geometric unification, mainly because the claimed “ppm agreement” is often not assessed against experimental uncertainties and because several integer/constant choices function like hidden degrees of freedom.”
Thanks for running this on GPT 5.2. It is fascinating to see AI critiquing AI-assisted work.
The critique regarding hidden degrees of freedom is a fair point. However, in curve-fitting, parameters are continuous: one can choose 4.1 or 3.9 to make the data fit. In this model, parameters are topological invariants (integers like 4 faces, 12 vertices, 20 faces). They are discrete and cannot be tuned.
The fact that this unadjustable logic yields results agreeing with experimental data within ppm implies either a massive statistical coincidence or a structural aspect.
It would be very interesting to run independent tests on different AIs with the whole context of the model and a standardized, consensual prompt. Beyond formal verification, this methodology could open paths that are difficult to navigate without AI assistance, helping to determine if the model stands as a possible foundation for a 'broad explanation of the observable', since the term 'ToE' instantly raises red flags. Kind of a pioneer peer-centaur-review. Just an idea.
I think a succinct way to describe my thoughts on linear algebra/language is that language has high dimensionality (ie many different basis vectors that may not necessarily be orthogonal) and that individual languages use a unique coordinate system to express thought. Each language is a lossy approximation of all conceivable thought and some languages can more efficiently represent the “all thoughts” vector space because they have basis vectors that point in more uncommon directions (like the go to japan example). So while you can more or less point to any thought in any language, some thoughts are easier to express in certain languages, which the post (and me) agree to be untranslatable words.
I tried to find the really interesting article about language and color that describes how some cultures use different naming schemes for colors but couldn’t find it. It talked about how back in the day we don’t know orange as a color, we just thought it was red-yellow and only after the fruit was distributed did the word for the color catch on. Here’s the best article I can find that talks about this phenomena https://burnaway.org/magazine/blue-language-visual-perceptio...
Each language is a lossy approximation of all conceivable thought...
This ultimately boils down to the private language discussion started by Wittgenstein. If you admit public language is a lossy approximation of meaning, you're taking a position on the existence of private languages.
> Each language is a lossy approximation of all conceivable thought
I'm not quite sure I understand this—I do have mental sensations/processes sans language, but I would not characterize them as "thoughts". To me, a thought is inherently linguistic, even if they relate to non-linguistic mental processes. So to me, learning a new language is very literally learning how to think differently.
I think we’re in agreement, but I’m afraid I don’t have the philosophical language to precisely pin my mental model into words (what a meta conundrum lol). I’ll try my best here, but I may come back in a few days with an edit if I can more coherently write my ideas.
I take a slightly more narrow definition of “thoughts” that may be more akin to “expressions” - ideas that can be communicated, so excluding non-linguistic mental processes. I think that may be where we disconnect. A lot of my idea about thoughts comes from the Borges story, Funes the memorius (short story about a dude who could not forget - interesting read and really clarifies my feelings on my definition of “all possible thought”). In the story he talks about tree leaves, but instead imagine needing a unique linguistic scheme for every single unique snowflake you ever see. It would be a linguistic nightmare! Therefore language must generalize otherwise it becomes noncommunicable and that generalization to me induces the “lossy approximation” I attribute to language in my prior comment.
So, in my head Funes’s mind represent the abstract space of all possible thoughts. When we use language, we are stacking words/sentences/paragraphs/etc together almost like vector addition trying to reach a particular point in the thought vector space. Some languages have really clean ways of getting to certain thoughts while others take a mouthful and still don’t get you exactly there (物の哀れ example from link).
I agree with your statement on new languages being different thinking. As you follow that vector addition process to get to the “thought,” different languages will take you on different paths to get to your destination thought because languages encode those vectors differently, even if the destination thought is the same. In my mental model, the act of thinking is putting those language vectors together and tracing their path to get to your thought.
And if my comment still makes no sense - I might have to incubate this thought a bit more :) but I do recommend the story- it’s a quick, thought provoking read.
> I take a slightly more narrow definition of “thoughts” that may be more akin to “expressions” - ideas that can be communicated, so excluding non-linguistic mental processes.
I was glad to read this because it seemed too neat and tidy for "thought" to necessarily be able to be encoded into language, especially in the presence of frequent miscommunication between people that share language, culture, and context.
On language and thinking, I agree that new languages promote thinking differently. But it seems that the difference has to fall short of the Sapir-Whorf hypothesis of informing perception or experience. Which would then limit the extent to which thought, as informed by language, would influence the way one would compose a linguistic representation of some thought/idea/"blob of meaning to be communicated." All to suggest that there is a broader landscape of "thinking-like activities" than those which would be able to be encoded linguistically.
Maybe it's simpler to say that I think of language as more lossy than thought.
I would argue that you can consider those thoughts. But this is the difficult bit, I've had the experience before of thoughts/feelings whatever ypu want to call them where words fall short. Knowing multiple languages helps a bit but it still falls short sometimes (very rarely).
Language is very effective at this, but I don't think thought is inherently linguistic.
To me language is just a way to label, group or organise these things. So when you learn a new one you learn a new 'labeling system/taxonomy' does that sound familiar?
I tried the new battlefield game and it’s bizarre how some of my friends play it. There’s this expansive battle pass (pay $20/quarter for access, on top of $70 base game) where you’re given tasks in game to make progress towards cosmetics. My friends only play to complete these tasks - the actual combat gameplay loop almost feels secondary to their enjoyment of the game. The modern trend of games becoming chore simulators is worrisome to me because -call me old fashioned- I believe the core gameplay loop should be fun and addictive, not the progression scaffolded on, even if that gameplay loops is a little “distasteful” like GTA.
It has to do with how ATC needs to be able to communicate with all planes in the air, even ones built 100 years ago. They have to use radio so everyone can hear everyone else. There’s no other technology that is as ubiquitous as radio, so they have to work with what they’ve got. Upgrading to other stuff would be an absolute nightmare, though they are making progress on less critical fronts.
Couldn't comms broadcast in multiple parallel modes, like cell phone traffic?: More clear (probably digital) transmissions in on band, and for backward compatibility, old radio transmissions in another.
I think one of the best things they could add would be an electronic drawing tablet for ATC to draw a flight path on a map and pipe it directly into the pilots EFIS or HUD. It's not fool-proof, but in high density airspace, it seems more efficient to be able to draw a curve and press a button than try to verbally describe it. Of course one major pitfall is you cannot draw in 3D.
and in this case, the pilot loaded the wrong waypoint (likely from another runway) and flew toward it. It's not slower, either. have you ever entered a waypoint into a flight computer? they aren't exactly built for speed.
Making the traffic controller load a flight plan into the plane’s computer merely changes who will make the mistake. If you want to suggest changes, suggest something that actually has a chance of reducing the rate of errors.
Besides, pilots don’t just blindly follow the waypoints on their computer. The pilots would have spent half an hour before they even got in the airplane reviewing the plan for the flight. This includes reviewing the published departure information for the airport. In that briefing they would have specifically noted that the direction they will turn after takeoff depends on which runway the tower tells them to take off from. They cannot necessarily guess which runway that’ll be in advance.
They already have something better than that. It's called a published departure procedure that pilots are supposed to follow. In this case, one of the pilots failed to follow the published departure procedure and came close to being on the next season of Mayday: Air Disaster.
The paths would be repetitive. Wouldn't it be great if, instead of drawing a new path in 3D space every single time someone files a flight plan, someone studied the area surrounding the airport, taking into account obstacles, traffic, the fact that there's a residential neighborhood on one end of the airport that shouldn't be bothered at night, and prevailing wind patterns, and drew all those 3D paths through the sky and published them for everyone to use, so that traffic follows known predictable paths?
And in the case of a deviation, would it be faster to pick up a stylus, draw a new path in 3D space, click send, and wait for a message acknowledging that... or someone yelling "27 alpha turn right heading 270 immediately"?
I understand the sentiment about a skilled user not needing this, but I think having a little buddy that I can use to offload some menial tasks would be helpful for me to iterate through my models more efficiently; even if the AI is not perfect. As a highly skilled excel user, I admit the software has terrible ergonomics. It would be a productivity boon for me if an AI can help me stay focused on model design vs model implementation.
I bet the folks who implemented the system do have checks and balances. The article said they placed 2 million successful orders which realistically can’t happen without some form of error correction. These reports seem like black-swan Taco Bell orders that break the system despite any safeguards against it. Luckily there’s no way the guy behind the counter is pouring 18,000 waters lol. I agree with you too - “Taco Bell Employee Fucks Up Order” is only newsworthy because an AI did it when the real headline should read “AI Successfully Processes 2M Taco Bell Orders”
If a gun manufacturer advertised a claim of "two millions bullets successfully fired", you might naturally wonder how many of them struck their intended targets.
I think the same omission is more telling than not, here.
"Our system cannot handle edge cases, but look at how many times we can walk the happy path" is not exactly an endorsement that would get it through QA on any team I've been on.
He mentions s&dbox, his new engine/gmod spiritual successor. They maintain an interesting devblog over at https://sbox.game/news if you are interested in how the sausage is made.
It feels like his new goal is directly challenging the niche that Roblox sits in. They have their work cut out for them if they hope to pull people away from that ecosystem.
The other https://sandbox.game have been trying to do exactly that for about 7 years and haven't managed it with hundreds of millions of investment and many large IP partnerships. Roblox and Fortnite are hard to fight.
Crypto/NFTs are a means to an end for Sandbox. They were used in the hope that it would make things better, not worse, by allowing users to freely sell their in-game assets on the open market. Unlike, say, Fortnite, where if you get banned all your assets are burned too.
I know it's a terrible argument but just don't get banned? Anyone that cares about the financial aspects of their assets shouldn't be doing things that could get themselves banned. Sure, false bans can happen, but is going crypto/NFT the solution?
Honestly, I like that the assets are freely tradeable on the open market. It's kind of a fun addition that you can on OpenSea and buy a new character to play with, or buy a bunch of buildings or NPCs to drop into your game.
I still think it was a good idea in theory. Crypto is a pain, though. It adds an annoying step to everything.
Probably a bad idea to use that name considering https://sandbox.game have been around for over a decade now, and despite this week's management shake-up they still have about $300m in funds in the bank.
Figured I’d ask the HN crowd- what’s the best way around these geofence blocks? Have you had success with a system that can work smoothly on mobile/desktop without any of the disastrous privacy and performance implications that VPN services are prone to?
Just use a lightweight, privacy focused VPN like Mullvad. You don’t have to keep it on when you need top network performance. Ultimately, a VPN of SOME kind is the only option.
I mean anything that circumvents geo-fencing at the IP level is going to be tunneling your traffic through an address that isn't blocked. Your options are all VPNs, the choice is only who's running it—you via a VPS or similar, friends/family/community, a service as you describe, or the public with Tor.
The friends/family option is probably the most broadly effective at circumventing the block since you'll have a residential address but at the cost of a lot of latency and bandwidth. The most performant option will be VPS services but lots of sites will block them as well out of an abundance of caution.
The black hole has two conceptual parts - the event horizon and the singularity. The event horizon is a one-way imaginary shell where once you pass it, you will end up at the singularity which is a point at the center of the event horizon. It’s the hole in black hole. Because the radius of the spherical horizon grows linearly with mass, but the size of the hole is fixed at effectively 0, it allows for a bit of sightseeing on your way to impending doom if the mass of the hole is large enough.
Yup, you’re trapped, so is light, and as gravity bends you and everything around you into pretzels, you’ll see everything yet nothing, as even the light will escape your retinas, before they pop like little grapes.
Eventually your atoms will make their way to the center singularity.
One of the more mind bending aspects of this is how the horizon becomes inescapable. The singularity is the only “forward” that exists anymore. You cannot conceivably go anywhere else. Every direction becomes “in”.
One could say the same thing about death (or life). Once you’re born, death is the only “forward” that exists. You can’t calculate its exact distance but it’s inevitable.
And seeing as how time is only something observed at the macro level and is still completely unexplainable scientifically, you're really hitting the nail on the head here.
It's not similar to this at all. There is still a safe direction which exists - if you could reverse your fall, it would take you back to the plane. There is no reversing your "fall" into a singularity. "Out" no longer exists. Even if you reverse your direction, you'll still be falling towards the event horizon.
It’s kinda like why they make compsci majors take a history class - the actual knowledge may not be particularly useful unto itself, but learning that information/skill tickles a part of the brain you dont normally use. So that later, when you need to use your whole brain to solve some problem, all parts of it are strong.
So are the skills themselves important? Not per se, but they represent “meta skills” that we want kids to develop when they are most apt to learn.
Ultimately, I think it’s a big TBD - a child’s mind innately wants to learn, it’s just unclear what kids are learning when exposed to so much tech early on and that’s what has teachers worried. I bet the kids will turn out fine, but we as a society won’t really know until the kids are old enough to tell us firsthand.
“As written, this looks closer to sophisticated curve-fitting (numerology with constraints) than a legitimate geometric unification, mainly because the claimed “ppm agreement” is often not assessed against experimental uncertainties and because several integer/constant choices function like hidden degrees of freedom.”
Thank you for sharing and happy holidays!