Much how we have mostly evolved away from physical labor and now need to go to the gym to maintain fitness. I imagine LLMs will do the same for our cognitive acuity.
These kind of startup excite me. More so than the "next killer app" or "look what we shoved AI into."
Cheap electric transportation can finally offset USA's lacking rail commutes and attack Europe's short flight market (which is already cheap). Now we need bigger planes or more frequent planes and the next startup to solve the energy needs.
> Cheap electric transportation can finally offset USA's lacking rail commutes
I'm not sure how building an electric aircraft would help with this. Are people going to suddenly switch to commuting by plane?
Fixing America's commute problem is primarily an urban planning one (ie, putting all the homes nowhere near anything) and not an insufficient-electric-aircraft one.
I don't think it's the daily work commute being solved. More weekly meeting in Cleveland for someone who works in Chicago. Which today is a 5h one way drive and a hotel stay.
Could you imagine the (rather justified?) NIMBYism for all the airports required for commuter flight? If rail can’t be done, flight definitely won’t fly.
I happened to be in Plattsburgh the day of this flight, on my way to a fishing trip on the lake. They used that facility for the long former SAC base runway, but I think it’s also appropriate given the nature of the NY North country that an ultra low operating cost aircraft would see flight there.
The death of most industry really hurt these regions, which are largely cut off from the prosperity of the normal economy. The ability to affordably link places like Plattsburgh or Watertown NY to the broader world. That region in particular prospered with rail links driven by iron and timber.
90 minute flights to places like NYC metro or Boston Metro would be transformative, and may even create new airline operating models.
Yeah… Fair point! I’ve always perceived NIMBY as “we want the benefits but we want someone else to deal with the geographic price.” I bet we could find people who legitimately want commuter airports but just not near their house… wait… how would that even work?!
It's clearly marketed for remote communities with little to no connecting infrastructure. Think Alaskan wilderness or Hawaiian island hopping. This isn't for getting from LA to Denver for skiing.
Even Ryanair turnarounds take like 30 minutes. A bit more than that and you can charge the plane comfortably. Fueling bigger planes is quite the ceremony.
Making the battery replaceable by heavy equipment would drop the turnaround time for these. The planes fly regular routes (mostly), and could store a few extra batteries at their regular destinations.
Odd added benefit: the potential ability to jettison a huge part of the weight in extreme emergencies to increase flying time by the backup ICE system.
Not enough population density with single family homes. And if given a choice between living in a condo + better urban planning versus suburban housing, well. And then there is NY.
You can fix the density problem, at least on one end, with park-and-ride lots; every successful commuter / regional[3] rail service puts parking next to the suburban train stations. You just need a dense corridor for people to go to.
For people commuting from suburb to suburb you put bus connections at the train station. Buses are lower capacity but cheaper to purchase and operate so you can run a bunch of them in low density[0] for improved coverage. If demand is really bad[1] you create a "microtransit zone" - i.e. offer shuttle taxi service to the rider's final destination.
What I'm describing is basically how UTA's FrontRunner commuter rail service works; but it's not materially different from, say, the NY MTA's LIRR or MNR besides the larger scale of the NY systems and the fact that New York is weirdly bad at buses. For various reasons America has actually done decently well at retaining the commuter rail systems that worked and rebuilding the ones that got scrapped. This is primarily because these systems tend to be small-scale enough that state & local governments can fund them with their own money and the meager amount of FTA funds available.
The comment you replied to confused "rail commute" with "commuter rail"; but the grandparent was specifically talking about intercity travel. Here is where America is actually malincompetent[2] at building anything. Our primary intercity rail operator, Amtrak, was built as a last-ditch service preservation measure as railroads were bleeding money on passenger traffic. It doesn't own most of the rail it operates on, and that rail has been actively getting worse as Class I railroads have been destroying their own infrastructure (especially Union Pacific). The two-and-a-half attempts at building high-speed services in the US have all been various flavors of bad:
1. The most successful, the Acela, works primarily because Amtrak owns most of the corridor. Even then, it barely gets up to speed because "high speed[4]" (80+ mph) trains have to be grade-separated in the US, which is expensive, and the Acela was sold on the promise of incremental upgrades that haven't really panned out
2. BrightLine partnered with a freight railroad that wanted to get back into the passenger market, and it's actually operating at "higher speed[4]", but it's also bleeding money. They have a second route planned from Los Angeles to Las Vegas, but it's a project they bought out because it was... also bleeding money.
3. California wanted to build their own Acela, with blackjack and hookers, and wound up massively overselling it to voters. At least in terms of how much it would cost and how quickly it could be built. They bet the farm on FTA funds that got impounded the moment Trump retook office. I have no idea if this will ever happen.
You'll notice I never mentioned population density, and that's because it doesn't matter for intercity travel at all. The vast majority of flyers get a taxi to the airport and a rental car at their destination, and it still works out anyway. The only difference to those kinds of travelers between a train and a plane is time and whether or not they have to take their shoes off before boarding the vehicle.
[0] Or, if you're New York, you run way too many of them through medium and high density areas that really SHOULD have a subway but don't anyway build QueensLink and NO WAY WITH QUEENSWAY
[1] From a customer perspective, anything less frequent than an every 30 minutes bus probably could be served better with on-demand taxis.
[2] Malicious? Incompetent? Why not both!
[3] I am going to use "commuter rail" to refer to both commuter and regional rail systems as they are roughly the same scale. The difference comes down to scheduling patterns: commuter rail is scheduled around 9-to-5 rush hour traffic while regional rail is built to be more broadly useful.
[4] Strictly speaking, "high speed rail" means 125mph or higher and anything less is conventional speed. Rail just was already pretty fast even before Japan decided to make really fast trains.
I think you could make a case for that, especially when landing in water. Quite a few dense urban areas people might need to get to have water nearby suitable for landing seaplanes. I think you would end up creating communities further from the urban areas dedicated to this mode of transport. It’s definitely not a simple or guaranteed idea, but it is actually feasible politically and technically.
Btw, I kind of changed my opinion on AI. Finally, everybody can build an app, so we no longer have to pretend the people who built one and made it big are some sort tech pioneers pushing humanity forward.
Likewise, if you're a nerd, you can spend your free time building stuff you thought you would never get around to.
Yes, AI as marketed is a bit of a snakeoil, but much more real and meaningful than crypto and social media. And math nerds finally can get a high paying job that's not about peddling ads or figuring out how to make a hedge fund even richer.
USA's lacking rail commutes come down to the fact that America's leaders made the deliberate decision to destroy its own rail network and force people onto planes. The electric range of this aircraft is 125 miles[0], which is at the low end of the range where conventional intercity rail is time-competitive with the plane. A bullet train would absolutely smoke this.
tl;dr I have a few qualms with this app[1].
Now, this thing might take off (pun intended) anyway, for one simple reason: a theoretically faster train does not compete with it if it doesn't exist. The midwest and west are chock full of towns whose rail connections were either abandoned or never existed in the first place, but have regional airports. This thing would be perfect to service feeder flights from hub airports to regionals or as a charter plane.
That being said I wouldn't expect this to revolutionize air travel. You're still going to be packed into a sardine can and made to pay extra for the privilege of bringing a reasonable amount of toiletries with you. And all the long-haul flights are still going to be using fossil fuels because the energy density of batteries sucks. But still, I'd rather have this than 10 million AI companies.
[0] 500 miles for hybrid operation on SAF, because the FAA would not certify a plane with such a small range
I'm not sure the Midwest is the intended target, here. I think it's more suited to small, frequent trips like island hopping or trips to remote settlements.
I'm rooting for electric planes not because I prefer them over rail but bc I share your disappointment and pessimism about American railway. I much rather take a train but when Amtrak NYC<->DC costs as more than a plane, leaves from the same destination, and takes hours longer then flying becomes preferable. There are many route similar to this. Only legup that Amtrak has is frequency. If planes become cheaper hopefully airlines will afford more of them and offer more frequent flights. I dream of a constant stream of aircraft flying between destinations. But never underestimate greed, airlines will probably reap the cut costs and charge customer the same as before :/
> USA's lacking rail commutes come down to the fact that America's leaders made the deliberate decision to destroy its own rail network and force people onto planes
This is utterly 100% false!
The US made the decision to focus rail roads on freight. You never think about freight because you never use it for anything personal, but it is still there. The US beats any place in Europe for Freight and few countries even come close to US for freight rail.
There was a decision made to get rid of passenger rail. However most of those destroyed routes are routes that nobody sane would put passenger rail on today even in high rail mode share countries. There are only a small number of routes should return - for most of the lost routes a bus is better. That isn't to say passenger rail is bad. However the routes we had were bad.
No, the freights are NOT OK. US freight railways are responsible for the "precision scheduled railroading" brainworm that makes freight genuinely worse for anything that can be carried by truck and isn't good for anyone:
1. Freight customers spend longer waiting for their deliveries, and pay more money for them. As a result, a lot of mid-size freight customers have moved to trucking instead, and the market has shrunk.
2. To save money as less-valuable cargo is pruned from their networks, freight railroads have been downgrading their infrastructure from double- to single-track with sidings. This would be acceptable, except...
3. Freight trains have gotten longer and longer in order to reduce the crew needed to attend to a given amount of freight. Sidings on single-track routes are rarely long enough to accommodate mainline freight trains, so passenger rail gets delayed, and grade crossings stay closed for way longer.
That last bit is actually why I personally complain about freight railroads. I am someone who has spent a good chunk of his 2026 lobbying Utah's state representatives to adopt the Rio Grande Plan[0]. You see, Salt Lake City is divided in thirds by UP freight mainlines, which get hourly freight through-traffic that paralyzes car and pedestrian traffic across the various sides of the city. The railroads themselves don't inconvenience passenger train traffic[1], but they absolutely make it miserable to walk, bike, or drive across the city, and we can't run any transit services parallel to the railroad for obvious reasons. Everything that has to cross the tracks reliably has to go over one of two rail-crossing bridges in the entirety of downtown.
The reason why you are saying "a bus would be better" is primarily a matter of federal infrastructure spending. Buses get to ride on the Interstate Highway System for cheap, and government-provided highway infrastructure is significantly better for both freight and passenger traffic than the privately run railroads can offer. But at the same time, buses are a demonstrably worse answer to the question of "how do we get a bunch of people from one city to another" than trains or planes.
Personally, if money was no object, I'd call for the government to run intercity rail corridors along highway rights of way with open interconnection, overhead catenary, passing tracks, and passenger-priority dispatching. These would specifically be open to both passenger and freight rail, specifically because both are suffering under the weight of hedge funds that moonlight as Class I freight railways.
[0] The Rio Grande Plan is a citizen-led initiative that calls for building a train box in downtown Salt Lake City along 500 West and burying all heavy rail freight and passenger traffic inside of it.
[1] ...because we built our own railroad parallel to the UP mainline. Amtrak's California Zephyr service is absolutely inconvenienced by Union Pacific, though - in fact, it's the worst delayed route in the entire network!
Best in the world is fully compatible with sick in several ways such as you named. They really should improve. However the results of what they have done are still really good...
A bus is better because you still need roads for last mile freight, trades and other things that can't run on transit. May as well also run a bus when those roads are not congested anyway.
Realistically, you can't replace the entire global fleet on any reasonable timescale. And when climate change starts being such a problem that even politicians can't ignore it anymore, I think we'll have much bigger fish to fry than building electric planes. Flying is a luxury, after all.
Now, I think this is very neat, but I doubt we'll ever set foot on an electric plane this century.
I have anecdotal experience that makes the ~$1,600 per house hold feel like an underestimate.
We purchased a Subaru Outback right around the first round of tarrifs. We were in the market for the base trim, no bells & whistles. We ended up getting the highest trim bc all the remaining base trims were post-tarrifs and and cost $3.5k more than the pre-tarrif higher trim that was already in the lot.
Mind you the Subaru outback was assembled in IL, USA so it is not just imported entirely from Japan.
And that's not considering all the other "consuming" we do (which is quite limited being outdoor ppl).
>We purchased a Subaru Outback right around the first round of tarrifs. We were in the market for the base trim, no bells & whistles. We ended up getting the highest trim bc all the remaining base trims were post-tarrifs and and cost $3.5k more than the pre-tarrif higher trim that was already in the lot.
The average car on the road is ~15 years old. The actual lifetime is likely higher. Therefore for proper tariff accounting, your subaru's contribution to this year's tariffs should be 3500/15 = $233, not $3.5k. Not to mention much of the items people buy are either:
1. domestically produced (this includes housing, the biggest cost for most people)
Gruez was spot on, our last car was 15 years old. Though I don't know if his stat considered resold cars. We wouldve kept it too but the work it needed was $8k. Parts were available and already imported, just that mechanics in the area had all raised their labor rate. I support liveable wages but that points to tarrifs having higher order effects than just the goods that they apply to.
There are three parts of this talk, the last part (~last 15mins) bluntly address the question "how human roles shift in a world with advanced AI?"
My main takeaways from the speaker's position are: purely technical skills will get devalued because they tend to be verifiable tasks and AI will get better. Also effort will shift from building to evaluation of AI and systems. Speaker also mentioned that more of our time will go to decision making while AI takes over delivering on decisions.
I found many arguments to be at odds with one another bc the speaker does not address how someone can gain understanding without accumulating purely technical skills. It simply does not add up that someone who lacks purely technical skills in an area can make good decisions. Experience is often a by product of accumulating technical acumine in a particular problem area.
"It simply does not add up that someone who lacks purely technical skills in an area can make good decisions."
I've been thinking about this and your related points, and I've been (with growing concern) instinctively in agreement.
Another analogy has been coming up. I started programming (as a kid) in the late 1970s, first with BASIC (of course) but pretty quickly moved to machine language for obvious performance reasons, and then to C, followed by many other languages. In the late 1980s, at university, I learned proper assembly language, and used it quite a bit.
Back then, I was pretty good with assembly/machine language. Not super star or expert level, but comfortably solid. Since then though I've spent very little time at that low level.
Granted, such early exposure and experience has been a pretty big fairly indirect benefit. But over the years, I've met and worked with many really smart, productive people who have never had any experience at that low level, and that lack was rarely noticeable.
For any given 'layer of abstraction', there is some use in being familiar with the lower layers. I think, in general, the closer the lower layer is, the more useful the knowledge is. For example, how useful is deep knowledge of CPU microcode to a Java programmer? Perhaps some, but not a lot.
In the tech world, GenAI is likely to push most everyone quite a ways up the tower of abstractions. We're in the middle of that right now, and it certainly feels pretty alarming to me.
The biggest uncertainty that comes into my mind is how insanely fast this change is coming upon us. My crawl up the abstraction tower happened over the course of about 12 or so years, from machine language/assembly to C then to Perl then to many other things. In retrospect, that felt pretty fast.
GenAI is seemingly pushing us in that direction many times more quickly.
The pipeline problem is already well-discussed. Just learn the stuff even though the AI can do it, so you can acquire that supposedly valuable judgment. My concern is that if the only things left are the subjective and the social, only people who talk a big game are going to be valued.
> Nobody has good answers for any of these things. It's a folly to think someone else has it figured out.
We don’t know that. Only history will tell us who figured it out and why.
What I don’t understand is why there is a clear decision between decisions and delivering on decision, other than the artificial need to retain control.
I don’t see a real roadblock that will prevent AI making better devisions than humans.
I always thought that the Gettier problem was an academic curiosity.
Someone who justifiably holds beliefs about the eventual impact of AI that later turn out to be true---can they be declared to have been knowledgeable about the subject matter?
It could be that the ratio of the amount of context required for a good decision to happen to the amount of context an entity is willing or able to provide is not high enough. E.g. the AI might dictate my life better than me if I upload every single thing about myself to it, but for privacy reasons I only tell it a small amount about me.
And maybe corporations make similar decisions.
(You might counter that there will be local AI. This is true but it’s still risky to let an autonomous system have all your sensitive information, especially in the AI hacking age.)
I was wondering that too. I suppose if we take it to the extreme end of the spectrum, "end world hunger" is a great decision but how to actually realize it is another problem.
Tell ya what. Go ahead and prompt your agent to build your next project completely from assembly. Why even bother with the higher level inefficiency since it'll be compiled away.
Report back how productive debugging is and the velocity of adding new features to a codebase you can't reason about.
Without strong coding fundamental, this is what it will feel like for the future of the profession.
Bun? Of sorts? Hard port to Rust in a week and then merged in? No way they reason about that entire thing.
A business won't die just because an LLM is the only thing that can reason about a particular code base. At least, not like if a company's own engineers couldn't reason about their own code 2 years ago.
I recognize you don't agree with it - our world is changing.
Giving up knowledge (password) is something that is typically scrutinized at the border as well. Had he just handed over the phone and the phone had abilities to self destruct if tampered with (e.g too many incorrect pin entries) -- well the gov's case wouldn't been much harder. If they seized property and accidently destroyed the data, then that's on them.
You are correct, you have to give up the phone but can't be compelled to give up the password, and you'd get it back some indeterminate amount of time later.
It's actually been on the books for a while (decades at least) that customs can search you at the border without a warrant even if you are a citizen.
This case seems to have become a big 'Trump bad' poster child (people are calling the US East Germany in these comments...), but if this exact scenario happened at least in the last two decades (I found an example upholding the searches from 2004) then it would at least be possible to charge them with deleting evidence. Even this probably would have been nothing if he refused to give up his password, not being required to provide a password has been upheld for years. They can seize your phone for some time but I'm unsure on the times they ask and then just let you move on when they find out your a citizen.
This is precisely why you should use a pin and not biometrics. You CAN be forced to use biometrics (“something you have”), but NOT a password (“something you know”)
reply