Hacker News new | past | comments | ask | show | jobs | submit | toss1's comments login

>>The systems “are headed the right way,” says Greg Brannon, the director of automotive research at AAA.

They may have some promising data, but from what I've seen driving 2019 and 2021 model vehicles from Ford and Mazda, they are not even close to ready for prime-time.

I find both vehicles consistently falsely trigger the red-flashing-and-loud-beeping "COLLISION ALERT!!" dashboard warning to both small pavement cracks/potholes, and also to vehicles parked on the outside of a curve. This has happened at least a dozen times in the past ~30,000 miles driven. If those were instead automatic braking events, they would have caused an unnecessary rear-end accident least three times, as if I'd surprise "brake-checked" the driver behind me.

Yet, in an ACTUAL near-collision situation a few weeks ago, driving at night ~50mph on a rural road, a car ran a red light right in front of me, and I had to full-on threshold brake, years of road-race training reflexes kicking in before I was even aware of it. I barely avoided a collision, stopping with smoking brakes and a passenger with a pulled back muscle from having been not quite square with the seatbelt, my front bumper about a meter from their driver's door.

The car never made a peep — it completely missed the incident.

Of course, a working automatic braking system might have helped anyone in that situation, including me if I'd been a bit more sleepy or distracted. BUT IT DID NOT EVEN DETECT IT. The track record for the past 30,000+ miles is:

— 100% false positives — 100% false negatives

These 'collision detection systems' are 2-5 years newer than the systems AAA tested, yet the track record is awful. IDK what they are smoking but I do not want any.

I doubt the automakers are somehow holding back some magic solution, and would give them a LOT more time to get it right.

Just because a technology is promising does not mean it is ready to provide a benefit released in the wild on fast-moving multi-ton vehicles.


Smoking brakes from a single emergency stop? Sounds like your car is defective and I'd get it checked out. Sounds like you were lucky this time and avoided the accident but get your brake pads changed and make sure the fluid is topped up as that isn't right. Potential issues are stuck calipers, worn out friction material, or boiling fluid.

With a well maintained car (i.e. brakes that don't end up smoking when doing just a single emergency stop, even from 60-80mph...) you are less likely to avoid accidents by just a meter...


It's a metaphor; I didn't actually smell any smoke (but didn't stay around either). But I guarantee those pads were momentarily very hot.

Thanks for your concern, but no, the car's braking capability was fully up to snuff and not compromised, stuck, worn-out, or boiling; it hauled the car down to zero in a handful of car-lengths, and mildly injured the passenger — crappy brakes can't do that.

And yes, perfectly good brakes, even race brakes, can be made to smoke in a single stop. It's done every time you take them out to bed them after install (many high-performance brake pads are 'transfer pads' that work best when some of their material is transferred at temperature to the rotor). Also, it is very common to see high-performance situations where the rotors are glowing-hot in a corner [0].

The reason it is so rare in ordinary street cars is most people not knowing how to use even 30% of the real braking capability. People either don't press hard enough, or just press mash the pedal and skid. Once the tire is skidding, the brakes aren't turning, so they aren't heating. Even ABS just splits the braking between the pads and the intermittently skidding tire.

Threshold braking is a skill that both takes time to learn and practice to maintain. It is basically getting a feel for the exact maximum braking grip the tire has at that moment, and braking just below that threshold. It extracts maximum performance from the car (sliding friction is always less than gripping friction) and also puts maximum heat into the pads.

And yes, threshold braking can really heat up the pads in one stop, and (depending on the car, track, and driver) completely burn down a set of pads in a single outing. OTOH, I had a set of pads last an entire race season on one car, or have to replace them every several sessions with other cars. Properly used, brake pads are just a wear item, but this is unfamiliar with street use, because you should typically almost never be threshold braking, so they'll last tens of thousands of miles.

But the main point here is that the ONE situation I've had in the last 30k miles with "collision alarms", the car completely missed it.

[0] https://www.youtube.com/watch?v=NBoMTYZAyJc

edit delete stray chars


Yup. Contrary to popular belief, large or dominant businesses like regulation, as long as it favors them. Beyond regulatory capture to explicitly write regulations favorably specifically to the capturing companies (as Musk is doing), just the presence of a large burden of bureaucratic paperwork regulations favors large established players; they have the scale and excess capacity to handle the overhead, where another few percent of nonsense overhead burdens can crush a marginal but growing competitor.

They do already, but do need reinforcement.

>>"“We put in place a National Maritime Information Centre in about 2010 and we needed a Joint Maritime Operations Coordination Centre alongside it, because we said very firmly we have to take threats to our territorial seas and exclusive economic zone very, very seriously.

They are now in place, which is good, but they need to be really reinforced and the departments involved need to fully man them, because otherwise we are not going to be able to counter what is a very real and present threat and could cause major major damage to our nation.”" [0]

[0] https://www.independent.co.uk/news/uk/home-news/uk-undersea-...


Substantial Russian activity also near UK, raises concerns that Russia would cut off UK. [0]

Russian ships ‘plotting sabotage in the North Sea’ [1]

[0] https://www.independent.co.uk/news/uk/home-news/uk-undersea-...

[1] https://www.independent.co.uk/news/world/europe/russia-ships...


The Irish just chased away a Russian "research" vessel.

https://www.theguardian.com/world/2024/nov/16/russian-spy-sh...


Interestingly, Ireland is not a NATO member, so it's somewhat surprising Russia is poking around there. Although they're still EU, so maybe that's why.

Are there not cables run through the Channel Tunnel? Seems like a no-brainer.

Yes, there is fiber infrastructure in the Channel Tunnel [0]. I'm pretty sure that while any one good link is vastly better than zero links, no one link is sufficient to carry all traffic from/to the British Isles?

[0] https://www.colt.net/resources/colt-successfully-completes-t...


If you're routing through the Chunnel, I suspect you could fit at least two seperate links.

I mean, the UK has 20+ fibre links to other lands. If one goes down, fine, if a second goes down, it's suspicious. If a third goes down, and there are Russian ships milling about over the location of the.. yes, there goes a fourth, it doesn't take long to realise what's going on.

Now, what the British Navy would do about this I'm not precisely sure. But even to escort the ships away would put a stop to it, and the UK wouldn't be cut off.


The silly thing is they know entirely well that we can do the same to them. The US/UK at least have at least the same capability, if not moreso.

Do these nations not have navies? Can’t they tell the Russian non-combat ships (or pressure them) to get lost?

We do but ocean and air is big :)

https://www.independent.co.uk/news/world/europe/nato-jets-in...

Also with cables, they can be destroyed with "innocent" ships that have a right to be there actually :)

https://www.bbc.com/news/world-europe-65309687

Russia has a programme to sabotage wind farms and communication cables in the North Sea, according to new allegations.

The details come from a joint investigation by public broadcasters in Denmark, Norway, Sweden and Finland.

It says Russia has a fleet of vessels disguised as fishing trawlers and research vessels in the North Sea.


> Can’t they tell the Russian non-combat ships (or pressure them) to get lost?

Not in international waters, which is where submarine cables are largely located.

And even if they could: The oceans are... kind of big. If it were that easy to "just patrol" shipping lanes/submarine cable tracks etc., why would piracy still be a concern?


Would relatively cheap AI-piloted satellite connected ships with sensor equipment work as a solution?

I doubt it. It seems to be a similar problem to missile defense: When you have a lot of ground to cover and can only be in one place at a time, the defender will always be at a huge cost disadvantage compared to the attacker. That's only in one/two dimensions – add a third (submarines) and the cost imbalance shifts even more.

And even if it works, this will only give attackers pause that are deterred by attribution.


Basically if mass produced something makes defense "cheap" it likely makes offense even cheaper.

Just because it is not publicized does not mean it is not happening. Most military operations do not take along journalists, and are not reported to the press. Some are even secret.

That said, there is a limited amount that can be done in international waters without creating an international incident. Law Of The Seas, Freedom Of Navigation, etc.. It is to our advantage for example, when we want to prevent CCP's from denying access to international waters around Taiwan or Phillipines, but to Russia's advantage when scouting undersea cables in international waters.

They can field more "research" vessels than we'd typically field mil vessels, but I'd bet real money that that ratio just changed a lot in the past few weeks, as it hits the press.


> They can field more "research" vessels than...

Back during the cold war, there was very often a Soviet "fishing boat" trailing after any substantial US Navy fleet. Said fishing boat may have had far more antennas than any fisherman would expect, but far less interest in catching fish.

Fast forward - what would be the cost of having cheap western drones hanging around nearby, when suspected Russian assets were close to undersea cables, pipelines, and such?


Fuel costs I suspect, which is where those continuous flight high altitude solar powered planes NASA was experimenting with really come into play.

That said, satellite tracking shipping is pretty easy - It's interdicting ina timely fashion which is not.


Agreed that interdicting - if that means a naval or coast guard ship, or a submarine - is far more difficult and expensive.

But cheap drones can transmit "don't do that!" warnings. And also video footage of the situation. Which would seriously change both the maritime law and political situations.


>Fast forward - what would be the cost of having cheap western drones hanging around nearby, when suspected Russian assets were close to undersea cables, pipelines, and such?

If the suspicion is high enough, it's pretty standard for a US submarine or surface group to shadow whatever it is. It's free practice for the submarine crew.

This happened when the Russian ships visited cuba earlier this year.


They do: https://www.theguardian.com/world/2024/nov/16/russian-spy-sh...

However, another one will be along soon.

I'd assume, at the moment, that the primary goal is intimidation rather than anything else.


Who says they don't do that constantly, but missed it this time?

And risk escalation!? /s

Seems the senior managers & senior workers/producers should be spending some of their time and resources to develop AI training systems including coursework, simulation rigs, coaching, etc. to accelerate and reduce the load of training entry-level workers. This would multiply the capacity of a fully staffed team.

If they lack the foresight to do this, then yes they will break the career ladder as well as the path to their own team's survival.


I'm not sure senior positions will be as immune to AI. Sure, the work of a sales manager will be less prone to automation. However if your junior headcount is divided by four because of entry level automation, there will be much less managers needed.

If it is an influence operation, the people who want to wield influence pay the bills. Already the point of X/Twitter (large Saudi funding, likely to help prevent another Arab spring type event in their country), and the point of the hundreds of millions SBF spread around. Bluesky's Series A was Blockchain Capital; seems like part of this year's significant movement of crypto influencers into politics. If so, they don't need it to turn a profit, they'll profit off the influence. Just like the corporations who normally jettison any money-losing department, but buy and keep permanently loss-making news departments for the influence they can create.

Paradox as old as recorded time:

Being ready for war prevents war : Being unready for war invites it.

It's really simple: If you are leading an expansionist state, who are you going to attack first? The neighboring village/fiefdom/nation that trains like Sparta and is clearly ready to kick your military's asses back to your farthest border, or the other one whose population is mostly too fat to run down the street and spends their time chasing the latest TV show and fashion trend?


This made me think if it would make sense to send SUV owners to battle in their SUVs. Might actually increase survival rates.

good thought, depends on the battlefield environment?

Most certainly.

Another thought further down the line is to get some self-driving equipment installed, and use the cars before sending actual people. I feel like a relatively cheap kit could turn one into a ram and/or remote surveillance point.


Installing a drive-by-wire kit on a car sure isn’t hard. Self-driving would be a bit trickier, need a sensor suite and some OSS like ROS2 or the like.

Spam is already at the level of changing people's default behavior from answering calls that come in to (as I saw described abt younger mobile users) "would rather pick up a live hand grenade than an unknown caller"

Personally, my default was to pick up and is now if recognized contact pickup, if not and I'm expecting an unknown call, scrutinize, then if pickup, only answer with a cough or two — never "hello" or "yes" due to threat of voice cloning escalate to banks.

Spam is universally detested

My assessment is that voice calls are on the verge of going obsolete if the telcos fail to get a handle on spam. Yet the telcos behave as if they have no clue whatsoever and DGAF.


>>we would need to use all the energy coming from fossil fuels to undo burning them

This would true if we need to re-create the original molecule with it's stored energy (plus losses of course).

However, it seems this is a misapprehension of the task. Instead of trying to recover the entire hydrocarbon molecule, we're "just" trying to extract or recombine the CO2 reactant.

Without doing the chemistry or the math, it seems likely that a variety of methods of either preferentially attracting CO2, or combining it into simpler lower-energy-dense molecules to be collected, would require less energy as was in the original hydrocarbon, often substantially less.

Seems it should be an inequality, not an equality. Or am I missing something?


While you are right that capturing the carbon dioxide can be done with relatively little energy, that is not what the article is about. If you capture it, you end up with tons and tons of waste, essentially as much as the fuel you burned, what are you going to do with it? The article is about [...] converting CO2 into useful products [...] so that you do not end up with waste but useful products and the requires as much energy as you got from burning the stuff, at least to a first approximation, you would of course not try to recreate the exact same stuff you just burned.

If you capture the carbon dioxide, then for every supertanker full of oil you burn you need to permanently get rid of a supertanker full of liquid carbon dioxide. This is of course a project of insane scope given that we burn billions of tons every year. So in order to not have to deal with the waste, what if we just turn it into something useful that people will pay for? Because that costs a lot of energy, the energy we just extract. And now you want to put it back in? To get back what you just burned or at least something similar that you could almost certainly produce more efficiently directly from the oil?


OK, yes, building larger-molecule more-useful-stuff will take more energy, and I'll go with the first approximation that it's a similar quantity of energy re-input (some useful things less, some more). And yes, all that product will take substantial volume. Thx for clarifying.

That said, it still seems an extremely useful measure, even if we keep using only single-digit percentages for long-use plastics instead of hydrocarbon fuels.

Let's assume that for the next century or so a bunch of applications will continue to require the convenience and energy-density of liquid hydrocarbons. In order to avoid extracting more and further increasing CO2 levels, we'll have to input significant energy to reconstitute them from CO2. Obviously, inputting that energy from more fossil fuels defeats the purpose, but using renewables will work; and now they are even cheaper energy inputs.

The result would be a cycle of newly fabricated hydrocarbon fuels, which can be custom-optimised for each application. No new CO2 would enter the atmosphere and the existing levels would be reduced by the amount of hydrocarbon fuels (and plastics, etc.) fabricated and in existence throughout the entire chain of existence, fabrication, storage, distribution, transport, in-vehicle, right up to the moment it is burned. With cheaper renewable energy inputs and optimized custom fabrication, it would likely get cheaper than the existing drill/pump/transport/refine process. And, it's permanently sustainable, and as liquid hydrocarbon fuel use declines, custom production can be converted to storable materials.


Totally agree, it makes sense to use renewable sources to produce hydrocarbons from the air, whether to burn them or for chemical products. But to significantly remove carbon from the atmosphere as suggested in the article it makes no sense.

I think they mentioned backfill radars were expensive.

Yes, it does, but it says nothing about the cost of any alternative.

If 1 farm is 5.5 terawatt hours per year then 13 farms is, what, at least 50TWhr?

Wikipedia tells me the biggest nuclear power plant in Sweden is Forsmark, which produces 3.3 GW and 14% of Sweden's electricity. That's 30TWhr.

So is the cost for a new radar system less than the difference between building and maintaining 13 wind farms vs. building a new reactor?

Oh, hey, I see there are contentious proposals to build new nuclear power plants in Sweden, with state-backed loans and minimum price guarantees, and this is a 'U-turn' in policy in just the last couple of years (quoting https://www.government.se/contentassets/69254b65d64f46fd866b... )

Now I can't help but wonder if this is cowardly way to tilt the balance in favor of nukes, by making it harder for any cheaper alternatives.


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

Search: