Posts by dvhamme

Danish techies claim they can predict your next move (and your last)

Nothing to see here for now, carry on

This is just the fancy version of correlation, without having to bother as much with having to actually select the variables or compute good features to correlate mortality with.

Your insurance company has been using a simplistic version of this forever to determine your premium.

Many commenters remark that their course of life is dominated by seemingly random events, but those random events are rare on the time scales presented here (4 years for the mortality prediction for example), and in terms of data space they represent a discontinuous change from one part of the space to a different part.

The scary part is that you could extend this to predict whether divorcing or marrying or quitting your job would be more likely to improve or worsen your survival rate... I remember a not very good movie about this with Ben Stiller (Along came Polly).

Boffins reckon Mars colony could survive with fewer than two dozen people

Pointlessly specific research

I can understand how simulations and models like this can help us understand group dynamics on Earth. Think "why does everyone end up hating this guy, and not the one with the clear asshole personality". But the methods are woefully inadequate for trying to accomplish what the authors set as their contributions:

- To understand social interactions of Mars base occupants: the personality types they chose are extremely generic, while no agency will send average joes up there.

- To consider planning challenges for a Mars colony: until we have solved the thousands of technical considerations, it is pointless to plan the social structure. Social life on Mars will be dominated by the technical constraints imposed on the occupants: space, facilities, schedules. Chain of command would be interesting to investigate, but they did not really go there.

- To propose minimum colony size: this is again dictated by technical requirements, which the authors guessed at. How much weight can you take there in one go and how long does it need to last.

In reality, any extra occupants over the first one that you need for bragging rights will be decided from a cost-benefit perspective, and then the social considerations will be solved "as well as possible" within the group, mostly by training and observing different crews on earth in actual, physical simulation environment. One guy won't do, if he gets sick, you are screwed. Two for redundancy. Three only if it's cheaper/safer (the two are largely exchangeable) to have a third high-maintenance low duty cycle low power universal tool (a human) rather than a set of highly specific robots / machines. There will not be an "indefinite" presence on Mars, it is going to be basically a space station like the ISS but in an extremely inconvenient location.

Re: "four possible psychological traits: neurotic, reactive, social, or agreeable"

Agreed, I think it would be far more useful if they had taken inspiration from astronaut programs and the personality traits they screen for, as I imagine the requirements are broadly similar for a space outpost as for a space shuttle. Intelligent, tough, calm and meticulous? I am guessing based on common sense since unlike the authors, I am not being paid to look it up.

Russia's orbital insanity is almost beyond redemption – but there's space for improvement

Should be a bit careful with the neurotoxin argument, as it is entirely clear once you delve into the Skripal case that the official British government narrative is a fabrication. I do not know what happened, but I am certain neither does the register or YetAnotherJoeBlow.

The truth is, honest people need willpower to cheat, while cheaters need it to be honest

Absolutely not! Investigating relations between biological differences and behavioral differences is hugely relevant, and this study may not be the best executed and the authors got carried away a bit with the morality of it all, but it is not a futile study.

Aviation regulator outlines fixes that will get the 737 MAX flying again

Re: Hmm.

I don't understand this line of reasoning. Every pilot in the world will now know that MCAS can cause pitch down when the AoA sensors pack up (which is now less likely as it uses two), it will only execute a single nose down input in such case, and the pilots will know how to disable it after that single event. The hole has been thoroughly plugged. In fact it was probably plugged as soon as the two accident reports achieved widespread circulation. The planes crashed because the pilots got confused and overloaded. If they had known what was happening, they would have executed runaway trim procedure and been fine. Perhaps they could even have left it on and fought it all the way to an emergency landing in reasonable safety, I bet if you put pilots in a simulator and give them a heads up that every x seconds the plane will nose down, they will be able to compensate and hold average altitude, and probably MCAS disables in landing conditions anyway.

So in my opinion there is no reason to avoid the 737MAX in particular. Either you say Boeing stinks and you boycott all their planes because who knows what other mistakes they made, or you now trust this model as much as any other airplane you don't know jack about (or you trust it more because of the scrutiny it is undergoing).

Segway to Heaven: Mega-hyped wonder-scooter that was going to remake city transport to cease production

Re: Image and price problems

I'm an avid cyclist in Belgium and on the few occasions I visited the US, it struck me that unlike in Europe, no one just gets on a bike and rides it in his regular clothes. And lycra definitely looks naff, even if I do wear it myself on long weekend rides.

Re: Image and price problems

There's a country of over 300 million people who still think bicycles are toys, and the ones that do ride them seem to go out of their way to look extra naff.

NASA to send Perseverance, a new trundle bot, and Ingenuity, the first interplanetary helicopter, to sniff out life on Mars in July

Re: Interplanetary Hellicopter

Gravity at this scale is very well understood. Perhaps it is sufficient to test the rotors and the lift they produce at various speeds in a partial vacuum representative of Martian atmosphere density. They can then simulate the lower gravity by testing the rotors and motors on a mock-up copter that is 2/5th the weight of the real thing which works out to the same lift requirement, but using good old Earth gravity. No surprises left in this type of physics.

Play stupid games, win stupid prizes: UK man gets 3 years for torching 4G phone mast over 5G fears

Re: It's a shame...

Agree with the sentiment that selecting content based on interests is a potential spiral of dumb, but it's very hard to implement a meaningful separation of interesting and dangerous content without falling into the pitfalls of censorship.

Beware, space Chuck Norris inside: Wacky flight rules for Chris Cassidy's first mission unearthed as Navy SEAL greets Dragon crew

Re: ISS vs LHC

Sounds a bit unfair, CERN grade boffins can still design the experiments to be performed on the ISS, and I don't think they'd be superior to current astronaut profiles in actually executing them. It's a bit like saying the guy cleaning the shit from the monkey cages has to have a PhD in biology.

You're 3 billion years too late to see Mars' impressive ring system. The next one will be along in 40 million years or so

Re: Serious question

I think in the general case, the spinning of the planet causes it to become an ellipsoid (it bulges at the equator and shrinks in pole-to-pole distance). This causes an asymmetrical gravitation field; debris are not pulled towards the exact center of gravity of the planet, but towards the equator because that bulge is closer. Thus all particles are gradually pulled into the equatorial plane, where they form a very thin disk.

What do you call megabucks Microsoft? No really, it's not a joke. El Reg needs you

"Redmond cloud and crud peddlers"

Short, powerful, fitting. Beige just really suits the corporate image I have of Microsoft.

Virgin Galactic takes another step towards blasting Richard Branson into space

Re: Priorities?

I don't consider myself a tree hugger but maybe it wouldn't be so bad if the world came to think of air travel as an expensive luxury again rather than a convenient way to quickly arrive at some ridiculously faraway place you perhaps didn't REALLY have to go to in the first place. At least Virgin Galactic is unapologetically elitist.

Academic showdown as boffins biff-baff over when Version 1.0 of Earth's magnetic core was released

Do we really need a paper for this?

The Rochester researchers started from ~3700 tiny samples, then whittled them down until they had 3 that showed something they wanted to see. Move along folks, nothing to see here.

Australia down for scheduled maintenance: No talking to Voyager 2 for 11 months

Re: So

I would hope they had a look at the most likely failure modes and concluded that just about all of them are either unfixable, or can still be fixed after 11 months? It's not like the thing is suddenly going to veer off course and need a human pilot to grab the stick and steady the bird... Even if it were, I don't want to imagine what Battlefield 1942 multiplayer with a ping of 16h feels like.

Let's just hope it's not going to gradually fill the outbox because Earth's mail server is failing to send delivery confirmation...

Over the Moon? Not quite: NASA boss has a good whinge about 'counterproductive' Authorization Bill

Re: Three steps forward and two steps back.

An ex-colleague of mine works at the Elisabeth station on Antarctica, which is limited in scale and scope but self-sustaining. He compared this to the American Antarctic village, which is huge and manned year-round and features an enormous pile of empty oil barrels as its most striking landmark. I bet they have ice cream too.

Remember when Europe’s entire Galileo satellite system fell over last summer? No you don’t. The official stats reveal it never happened

I think the article would be much improved if it focused on how unreasonably low a target uptime of 75% is, rather than claiming "high school level stat manipulation". Average uptime over the course of a month is what it is, nothing was manipulated here.

Basically the performance targets were designed so that one week of downtime EVERY month is acceptable, which clearly is not what the end users consider acceptable.

Found on Mars: Alien insects... or whatever the hell this smudge is supposed to be, anyway

Re: Maybe

I don't think NASA or any other space agency is taking a "Mars is dead" attitude. It's just that space agencies tend to be clued in on science, and science tell them that if there is life on Mars, our best bet to find it is to look for subtle biomarkers in geological samples and not by driving around looking for pyramids, towers or highways.

Remember the Uber self-driving car that killed a woman crossing the street? The AI had no clue about jaywalkers

Re: Simple solution

Why do we need to expect this? Why does a cost / benefit analysis not work in this case?

If you have a system with many components, and you identify the worst performing component (driver), and can replace it with something which is better averaged over all circumstances, why would you not do it, and instead demand the new component must surpass the old component's performance by orders of magnitude?

Re: Simple solution

That is ridiculous. Does a human driver go bankrupt when he accidentally runs over a pedestrian? No he doesn't, in most countries not even when you can prove that he was deliberately inattentive (drunk, catching pokemon).

The problem with autonomous vehicles is that we want them to be 100 times safer than the average human driver. That is totally unreasonable. If we would only demand that they make the road safer (i.e. they cause fewer accidents, or lower speed accidents than the average human), we would be zooming around in driverless cars RIGHT NOW.

Why don't we let the insurance companies sort it out? You buy an autonomous car from brand X. Your insurance company looks at accident records for this model, technology used etc. and makes you a quote which you accept as it is offset by the time you can spend doing other things in the car. If your autonomous vehicle then has a fender bender or kills someone, your insurance company pays for damages. Vehicle manufacturers have incentive to make their cars safer so they are cheaper to insure than the competition, we don't need entirely new accountability laws.

Maybe the idea is flawed, I just came up with it. Feel free to point that out to me.

I read the report and I disagree with the emphasis on jaywalking. The fatality is a result of 1) poor tracking code and 2) poor sensor fusion.

For 1) they should obviously consider associating a previously found "unknown" object with a freshly detected bicycle with overlapping or nearly overlapping position to generate a motion hypothesis.

For 2) at one point they assumed the bicycle was following the normal direction of travel in the lane it was detected in, but this hypothesis could easily have been proven wrong by asking the radar what it thought of it; the doppler radar would have picked up on the fact that the object had a near zero velocity along the viewing direction.

We don't need cars that explicitly consider jaywalking. It would have been sufficient if the system had been able to reconstruct, even with low accuracy, an estimated motion path and understood that this was abnormal for the road context. The system is designed to do that, but poor at it apparently.

Heads up from Internet of S*!# land: Best Buy's Insignia 'smart' home gear will become very dumb this Wednesday

Re: Contact your credit card company.

"With an avec" is the dumbest three word combination I have read this month. Granted we're only the sixth, but I doubt it will be beaten.

When one of NASA's sun-studying satellites went down, AI was there to fill in the gaps

Re: Neural nets

This can work but it's not as easy as you may think. You need a big, complex network and LOTS of training data if you want to get anything resembling real IR output. Basically your network must estimate the object boundaries, object type, object context and environment conditions to give you meaningful IR results. Those tasks in isolation are currently in the realm of the possible, but in combination I'm not convinced at all.

To focus the thoughts: a visible light pedestrian detection network (e.g. faster R-CNN) has hundreds of millions of coefficients, and researchers still put IR cameras in tandem with it because there is information in the IR that is simply too hard to infer from the visible light image.

I fully agree. The synthesized data itself is of no value, the value is in mapping the correlations and dependencies. But using a neural network to model the dependencies is basically cheating as you don't end up with a model, only a number spewing orifice. It's like Isaac Newton publishing a lookup table of falling times of spheres of different weights with no accompanying formula.

Consumer campaign to keep receiving printed till receipts looks like a good move – on paper

Re: email it to the company

Great idea, then they can mail tailored promotions disclosing your personal details to an address you don't own. "Dear Mr. TomPhan, get a great deal on these essential accessories for your Excelsior Epilator Bundle Including Tight Spot Adapter (Hot Pink). Limited time offer!"

Does the environmental argument really make sense?

Maybe all of the world's IT infrastructure could indeed account for 8% of total greenhouse emissions by 2025 (note that it is a prediction by greenies and I did not bother to check its accuracy). But that number is irrelevant, what matters is the environmental cost of a tiny sheet of paper versus the environmental cost of <1kb of data saved by the store and by your email provider.

A study by Carnegie Mellon came up with 7kWh/GB for storage and access of files in the cloud. Round up to 7e-3Wh or 25J per e-receipt (e-ceipt? Let's be hip!). A quick search yields numbers for paper production (without printing) of about 3MWh/ton. Low grade paper is something like 55g/m2. A medium sized receipt will then round up to 1g. Just the paper it's printed on is costing 3Wh or 11kJ. Someone should check this math (it's dangerously close to Friday) but that would suggest an e-ceipt is over 400 times more environmentally friendly than the paper that goes into a real receipt, to which you may add printing and disposal costs as well.

As halfwit, would-be dictator buried by UK judges, Spain would like to dig up a very real one

Re: Yes writers, you are fashionably left wing

Did you not read the dig at Boris Johnson in the headline?

NASA's lunar spy looks for hide-and-seek champ Vikram, Starliner test success, and more

Re: Eh?

I too had to look it up. https://www.youtube.com/watch?v=X5O77OV9_ek around 1:05 you see how the orbit closes in on the moon periodically and makes a sharpish turn around it every time. Very interesting.

Did you know? The FBI investigated Gamergate. Now you can read the agents' thrilling dossier

Re: Humanity fail.

Yes, bad things happen to women just because they are women. But what is worse, is that certain women misappropriate the sympathy that goes towards victims of these bad things for their own benefit, even if they are not actually victims. Zoe Quinn actively sought out bad reviews of her (universally accepted as bad) game which she then twisted into attacks on her person so she could stir up a media storm about her being "harassed". As a result her bad game got unwarranted coverage, and she became famous. I don't know how the other two fit into the story exactly, but it's not a stretch of the imagination that Wu running for congress is a direct result of this self-started vicious circle of hate.

I mean, I don't consider myself a misogynist. I never had anything against this woman (Quinn). But I do now, I hate her with a passion for what she represents. If I were a juvenile idiot, I'd be ready to harass her, and that'd play right into her cards.

Dark net LinkedIn sale looks like the real deal

I deleted my account when the breach first became public. When someone asks me if I'm on LinkedIn I gladly tell them why I'm not; it seems my reasons are still valid. They don't even take security serious after some hacker puts their amateurism on the front page.

Brits rattle tin for 'revolutionary' hydrogen-powered car

Genius business model: charge a reasonable price for lifetime usage. Because when the car fails, you don't need to replace it; most likely it was the hydrogen tank that went, and the driver is out of the picture anyway! So whether it explodes or keeps running, the house wins.