It would be in Google's ultimate interest to label AI-generated websites and potentially rank them lower in search results.
AI needs to be kept up to date with training data. But that same training data is now poisoned with AI hallucination. Labelling AI generated media helps reduce the amount of AI poison in the training set, and keeps the AI more useful.
It also simply undermines the quality of search, both for human users and for AI tool use.
Why is the toggle allowed at all? Presumably, sometimes we don't want apps to know where we are and record/share that to the highest bidder.
International travellers will know that some apps will alter behavior or refuse to work based on your location, if it's provided. If I use a VPN, I want the app or website to use only the IP location*, not the radio location.
Modern phones connect with a randomized MAC address. So yes, you can track a person around, but you will need another system (like the WiFi login page) to match MAC to identity.
I don't feel that this article is a fair summary of the paper. And the title is just clickbait.
The paper says, in a somewhat contrived scenario, with dozens of labelled walkthroughs per person, they can identify that person from their gait based on CSI and other WiFi information.
This is a long way from identifying one person in thousands or tens of thousands, being able to transfer identifying patterns among stations (the inference model is not usable with any other setup), etc.
All the talk of "images" and "perspectives" is journalistic fluffery. 2.4Ghz and 5Ghz wavelengths (12cm & 6cm) are too long to make anything a layperson call an "image" of a person.
What creepy thing could you actually do with this? Well, your neighbor could probably record this information and tell how many and which people are in your home, assuming that there is enough walking to do a gait analysis. They might be able to say with some certainty if someone new comes home.
That same neighbor could hide a camera and photograph your door, or sniff your WiFi and see what devices are active or run an IMSI catcher and surveil the entire neighborhood or join a corporate surveillance outfit like Ring. Using the CSI on your WiFi and a trained ML model is mostly cryptonerd imaginiation.
Indeed. I'm confused by this line from the article
> a study with 197 participants, the team could infer the identity of persons with almost 100% accuracy – independently of the perspective or their gait.
The paper seems to make it clear that the technique still depends on gait analysis, but claims it's more robust against gait variations.
It feels rather more than a little bit creepy to realize that Comcast et al, and thus the US government (if you live there), laundered through 3rd party data brokers, knows if you're sleeping and knows if you're awake. Knows if you've been bad or good, for ICE/ATF/DEA/SEC's sake.
Comcast is late to the party, then. AT&T has been selling your information for decades. And your mobile provider can track you anyplace that there's a cell-signal, potentially even outside the country.
The creepy part isn't the tech itself, it's that BFI data is unencrypted by default. So it's not even that someone has to hack your router. Any device in range can just passively read it.
While I don't think it would prevent our troops from having foreign-produced trucks in theater, we can't affordably procure such trucks thanks to the Chicken Tax. I would also guess that giving a DoD contract to Toyota for a truck that may not be registrable in the US would also face institutional resistance.
The military has an incentive to ensure there are plenty of Americans who know how to design and manufacture things. A truck and a tank have a lot in common - if war breaks out we want the ability to take people of of trucks and get them making things the military needs.
This is the same reason the Navy has for building ships in the US even though they can be done other places cheaper.
Maybe in 1942. Modern tanks cannot be built on highly specialized production lines that build road vehicles without years-long re-tooling. M1 Abrams tanks don't even use piston engines, they have turbines.
A older, but well documented example how specialized modern automotive production has become is the Mercedes Benz 500e. In the 90s Mercedes wanted to build a more powerful, wider version of the E class. They added 56 mm to the front fenders and discovered it wouldn't fit through the production line properly. MB contracted for Porsche to handle the low-volume 500e on a different production line.
> This is the same reason the Navy has for building ships in the US even though they can be done other places cheaper.
You'd think the biggest war machine on the planet would benefit from economies of scale by now. If they want to stay sharp they could build commercial ships between the ocassional war ship.
The company that builds aircraft carriers in the US tried exactly this back in the 1990s. It was a complete failure. They simply weren't at all competitive with foreign shipyards.
If you don't believe in the power and corruption of the military procurement industry and the military itself, then your comment is so unrealistic as to be deluded.
If you do believe in it, then it's simply irrelevant. Given the other reasons that the US military is spent with profligacy on US manufactured goods, maintaining 'truck know-how' does not register. If the know how consideration did not exist the money would still be spent in exactly the same way.
Also, because of CAFE standards, the US can't even attempt to create its own competing light trucks as everything needs to be fucking massive to maintain the emission exemptions.
The thinking was it would make cars more efficient but instead everyone just built obscenely large vehicles that were classified as trucks instead of passenger vehicles.
As much as I like to slag on CAFE, we have been here before.
Automakers simply hate making affordable cars. MBAs extol "Number must go up! BRRRRRRR!" and you cannot do that with cheap cars.
Remember the 70s? What did the big automakers do? They made bigger and bigger cars ever shittier and jacked up the prices. Sound familiar?
And then what happened? Japan showed up and cleaned their clock. And then the protectionist laws got passed, but it didn't matter because the Japanese cars were smaller and better and used less gas. Sound familiar?
History may not repeat itself, but it sure likes to rhyme.
There are two ways to improve fuel economy. The first is technology (fuel injection, aerodynamics, hybrids, etc.). The second is to make the vehicle smaller.
The first one is a trade off against cost, but the market is already pretty good at handling that one on its own. Fuel injection and aerodynamics don't add much to the cost of a car, so pretty much everything has that now. Hybrid batteries are more expensive, but the price is coming down, and as it does the percentage of hybrid cars is going up. You don't really need a law for this; people buy it when the fuel savings exceeds the cost of the technology.
The second one is a trade off against things like cargo capacity. If you say that "cars" have to get >35 MPG at the point before hybrids are cost effective, or keep raising the number as the technology improves, it's essentially just a ban on station wagons. And then what do the people who used to buy station wagons do instead? They buy SUVs.
The entire premise is dumb. If you want more efficient vehicles then do a carbon tax which gets refunded to the population as checks, and then let people buy whatever they want, but now the break even point for hybrids and electric cars makes it worth it for more people.
That's the thing though, they can't, at least until the very recent advent of EVs. We used to have similar vehicles (the old 80s/90s ford ranger, tacoma, etc) but they were regulated out of existence by CAFE standards.
Even if you repealed CAFE today, the automakers have all built their entire business strategy around selling enormous expensive vehicles and generally despise producing lower cost options.
We are starting to see what appears to be the beginnings of a small pickup renaissance due to electrification but none have actually hit the market yet and trump has further stalled that progress by messing with EV subsidies and environmental standards.
The current Hilux is extremely close in size to the US Tacoma... It has also grown over the years. Although if you look at the footprint size (e.g. what CAFE measures), the Tacoma has the wheels a little more advantageously placed.
I am sure they could consolidate the models to work in both the US and abroad, but my guess is they do enough US volume that it is not yet advantageous to do so. There's already a number of major parts that have been shared recently between the Tacoma and Hilux... e.g. the 2TR-FE engine and AC60 transmission. But usually Toyota chooses to spec the Tacoma as a more up-market vehicle, which makes sense given the US market.
In 1993 I paid a shade under $10k for a new Chevy S10 where the only options were AC (not actually optional in Texas) and CD player in the radio. It was manual transmission, V6. Indexed to inflation that would be, what, about $24k today if regs allowed them to be built?
If it existed they would fill every rural high school parking lot in the south. Allow them to exist and someone will build them.
My dad paid like $14k new for his pretty basic 2004 tacoma. It had the frame recall not very long ago but the bumpers are all rust pitted. Still doesn’t stop strangers from regularly giving him cold offers in the parking lot for close to what he paid new 20 years ago. People desperately want these trucks.
They are massive because of the cafe standards. There's plenty enough of a market for smaller trucks, even the Ford Maverick which is closer to a car with a bed sold out immediately.
I like my big truck but when it dies, if there's a small truck available that lets me plow snow and tow logs in the forest, I'll get it.
Smaller trucks don't have the suspension etc. for mounting a snow plow, or heavy towing. Those are applications where a bigger heavy duty truck make sense.
What percentage of drivers actually put their phone down when they start moving, would you say? Anybody will slowly normalize the behavior. It's just stopped traffic. It's a traffic jam, we're only moving 5mph. I'll put the phone down if we actually get going, etc.
You can just get some ball mounts and an action camera or two. The new cameras have such good stabilization that a handlebar mount is very acceptable now.
https://www.amazon.com/dp/B0BPMBL924 + a ball for the handlebar and a short extension. You can use the smaller ball size, it is still plenty sturdy.
> Is the end result just a couple of minutes later in a 30 minute commute?
More like a few seconds.
Every car that merges in front of you only costs you their following distance. If the average following distance is 1 second, then you are simply 1 second slower than you'd have otherwise been. So unless this is happening continuously every 30 seconds on your 30 minute commute, you will lose less than a minute.
The "but if I kept reasonable following distance, people will keep merging in front of me and I'll lose time" excuse is pretty thin given this analysis.
And an insurance claim can easily eat 40 hours of time between the insurance companies, other lawyers, buying a new car, medical appointments and recovery. That's 19,200 minutes you won't get back, or about 52 years of driving 1 minute slower each day.
Some amount of that is inevitable, but there is another level of defensive driving where you anticipate poor behavior and arrange that it won't cause an accident.
Have a look at a few dash cam accident videos [1]. There are many maladaptive patterns of behavior, but a frequent one that the average good driver can improve on is limiting speed on two occasions: when approaching a blind spot, and when passing stopped or slow traffic.
That second one gets lots of otherwise good drivers. They seem to think that by limiting their speed vs slow/stopped traffic they'd be encouraging people to dart in front of them. Which is somewhat true. But with limited speed, that's an avoidable or less injurious accident. By gunning it past stopped traffic, you make the accident unavoidable and more serious.
Without wanting to paint with too broad a brush, I would say in my experience driving in 10x countries, U.S. drivers, being most habituated to spending their lives in cars, drive in the most distracted, least careful way. Especially in places where the typology is the U.S. default of low-density, car-oriented sprawl. Accordingly there are an appaling number of deaths and injuries on the road: 1 in 43,750 people dies each year in the U.K. in automobile accidents vs. 1 in 8,500 in the U.S.A.
Inb4 deaths per mile driven, I'd argue higher VMT in the U.S.A. only proves the point - too many cars being driven too much because of silly land use. High VMT is acutally a symptom of a dangerous mobility system as much as a cause.
Add to that the relative ease to get a license in the US, and the level of punishment for breaking the laws.
I (an American) was on holiday and Switzerland and was explained the process of getting your license back if you lose it. It is a big disincentive to driving badly and putting yourself at risk of that, to be sure.
Adding on to this, a common reaction I see to online videos of driving incidents is "why did this person just stop? of course you are going to crash into them. They shouldn't have stopped" and many people agreeing with it. It seems they are blind to the fact that if the following driver was using a safe following distance and speed, they should easily be able to stop, making the incident the fault of the driver following too close, not the one stopping.
I haven't checked on this in a long time, but IIRC, the insurance company will always blame the person in back in a rear-end collision, for just this reason. A rear-end collision should always be avoidable.
Usually but not always. A common insurance scam is to pass a car, cut in just in front of it, then brake hard causing a collision. Dash cams video is a good thing to have to fight this if it happens to you.
No, even then. If a car cuts in front of you, you have at least a few seconds of time to start making space for them, and once they are in front of you, you should immediately make a safe amount of space between you and them.
Yes, in a country where safe driving is not internalized you will inevitable have someone rear-ending you while doing this, but if the options are "accident where you're at fault" and "accident where you're not at fault", pick the latter.
Some degree of road safety depends on predictable behavior. I haven’t seen those videos, but suddenly executing a panic stop on the freeway for no good reason at all increases everyone’s risk, even if the car behind you is following at a safe distance. Obviously the following driver bears the most responsibility, but erratic drivers shouldn’t be held to be morally blameless.
People don’t usually act erratically for no reason. Maybe they suddenly stop because they see a deer sprinting towards the road off in the distance, and the person behind them didn’t see it. There are tons of reasons that look like they “erratically stop”, which are actually genuine safe behavior that the other may not know about.
I learned next-level defensive driving by bicycle commuting to work 5.5 miles each way on busy roads in rush hour traffic. On a bicycle you're invisible, and if you expect any less, you're going to get hurt. As it was, I had some very very close calls- at least one of them had the potential to be fatal. Ironically, the only time I ever crashed was my own fault.
But now even when in a car, I retain that "I'm invisible" mentality, which makes me much more aware of what other drivers are doing, and much more skeptical of their ability to make good decisions. This has saved me several times.
The landing page video's first incident is a car coming from behind and from the right, cutting off the filming car. The filming car didn't react at all when instant (but measured) braking would've been safer to start building a distance buffer.
One thing HPDE taught me is that most people under brake in dangerous situations because they simply don't know the limit of their vehicle nor the sensitivity range of the brake pedal.
The hard braking heuristic makes sense when estimating risk of road segments, but not as a proxy for driver competence.
It certainly makes sense as a proxy for competence across a diverse population for insurance purposes. You have a baseline of hard braking events that a competent driver may encounter under normal circumstances. If a driver routinely exceeds that number, they are either unable to correctly estimate closing distance and reaction times, which makes them higher risk for causing accidents, or they are driving abnormally aggressively, which also makes them a higher risk for causing accidents. If you consistently put yourself in situations where hard braking is required, it doesn't matter what your skill is, you've reduced your safety margins and an accident is statistically more probable. You said it correctly with "would've been safer to start building a distance buffer", that is the proxy the insurance companies want to use for risk assessment.
35 years without an accident on my record isn't because I'm a magnificent driver, it's because I always try to leave a way out for when something unexpected happens, because the unexpected _does_ happen.
The fact that some people may have the skill to drive more aggressively means nothing in the aggregate as far as insurance companies are concerned. If you are skilled enough to drive in that manner, you are skilled enough to avoid it as well. It's simply statistics.
> You said it correctly with "would've been safer to start building a distance buffer", that is the proxy the insurance companies want to use for risk assessment.
Then use it? Mandate reaction speed tests or other driving mechanics competency evaluation (not road sign comprehension) and watch insurance margins explode.
The driver in my example did poorly and scored top marks in the heuristic.
> building a distance buffer", that is the proxy the insurance companies want to use for risk assessment.
The cam car did not need to have a hard braking event (HBE) to start building distance.
Even if they did, the insurance companies are looking for a pattern of HBEs to assess risk. I agree that there is a theoretical high-risk driver that never has a HBE because they always try to maneuver instead of breaking. There are other heuristics for this (high lateral acceleration, high jerk). And the ultimate heuristic: failing to avoid accidents, thus having a claim history.
> One thing HPDE taught me is that most people under brake in dangerous situations because they simply don't know the limit of their vehicle nor the sensitivity range of the brake pedal.
On any modern car, just push it all the way and let ABS and stability control figure it out, and don't let the vibrating brake pedal spook you into releasing it. That's just ABS doing its thing.
Really though, getting a license is too easy in the USA. We really need to require some sort of car control course, including obstacle avoidance in the rain. Would be really nice too if it included an obstacle avoidance course in two cars: A huge SUV or pick-up truck and a more reasonably sized sedan. So many drivers think they need a huge vehicle to be safe while being completely unaware of how well smaller cars can likely avoid the crashes to begin with. Would probably get really expensive really quickly, though.
One thing I noticed among dash cam videos is very often the person recording and publishing the video will keep on driving closely behind another car that is clearly driving erratically. Maybe he will honk, but he won't brake or leave any safety distance, and seconds later the idiot's car, which has been behaving weirdly in front of him the whole time, causes him to crash.
I realize this may come off as victim blaming, but I feel you should have an obligation to not endanger yourself even if by the laws of the road you are technically in the right. I would rather get cut off by and idiot and be at my destination thirty seconds later than having to deal with car repairs even when it is legally speaking not my fault.
> Early clock - keeps time anywhere between 0 and 10 minutes fast. For those who like to set their watch ahead to avoid being late. This clock keeps you from trying to "compensate," because you never know how early it is at the moment.
That's pretty genius for many ADHD-type folks. Only problem is a modern household has many clocks in view, so you'd need to commit to just not setting them.
Oh now that would be a fun version 2 challenge: have all the clocks in one household synchronize such that they're all early by the same amount at any given time.
Easy enough for wifi enabled ones: a UDP broadcast to discover other clocks on the network, then sync how you will.
For non-wifi-enabled clocks, perhaps something like a CH572 would do the trick: a $0.20 RISC-V microcontroller with BLE support that all the clocks in the same vicinity could use to talk to each other.
You could really mess with your neighbors if they had the same clocks and you were within range...
AI needs to be kept up to date with training data. But that same training data is now poisoned with AI hallucination. Labelling AI generated media helps reduce the amount of AI poison in the training set, and keeps the AI more useful.
It also simply undermines the quality of search, both for human users and for AI tool use.
reply