What on earth was that? Couldn't make head nor tail of it.
664 posts • joined 23 Jun 2009
What on earth was that? Couldn't make head nor tail of it.
I want to know what happens when the chef murders someone with a kitchen knife, while this robot is watching. How will the rampage be stopped?
> Strange, that's how they tech robots in things like car manufacturering. See how the human does it and then repeat.
1. Cars are made from parts which all tend to be exactly the same size when they arrive on the line.
2. Cars all tend to look exactly the same, with a few specifiable variations, when they leave the line.
So, if I want a red car with a CD player, you play programme "colour:255,0,0" at the spray stations, and "Stereo: 1" at the dashboard assembly station. Green with the DAB radio? "0,255,0", "Stereo: 2". Simple.
Now, let's say I want a carrot soup, but a bit more salt than last time, and a bit less pepper. And only for 2 people, not 4. Remember, this robot doesn't weigh the salt and pepper, it just shakes the salt and pepper pots exactly as it's been shown how to. Nobody has ever shown it how to make soup for 2 people, or how long to shake the salt and pepper for. And the carrots are a bit smaller this time, too, and one of them is forked.
A robot taught in this way will fail, spectacularly, at this style of task. Sound like I'm being silly? I'm really not - the process of machine learning they have suggested might work for extremely repetitive, predictable tasks, but is absolutely inappropriate for a chef.
A sensible design for a chef robot would be a big box, with tubes and pipes going to different locations, and as few moving parts as possible. It would look very like an industrial food processing plant, only minaturised. Maybe it could be in a small 20' container, and a conveyor belt could bring the finished products to a hatch? Designing a humanoid to do this task is both silly, and many, many years away from being possible.
Actually I suspect a constitutional crisis consisting of the monarch attempting to sack the government would, in fact, impact our daily lives quite significantly. Who knows how - most likely there would be a very severe recession (perhaps even making the pending Brexit mess look like a cake walk). It would certainly be more, uh, entertaining than "Does the media have a right to see Prince Charles' letters," for example.
I've always thought "Wallops" is a great name for a space port. Sadly missing the h though.
There should be another called "Wacks".
We live in an area where there should be hedgehogs, but aren't. We'd really like some in the garden.
Maybe I should try the Facebook market?
> My life is enjoyable because I interact with humans and nature, not a screen.
...he typed, into a screen, on a tech website... I think you forgot your
The world needs early adopters. Sometimes they get burned, sometimes they don't. We've all been there, and if this guy is enjoying his early adoption, more power to him. If it means the price comes down in a few years for the true mass market, great. If not, fine, he's enjoying some unique hardware.
Really unsure why he is being called a fool, and downvoted, for that.
> Unless you're referring to the bluetooth interface?
Of course he's referring to the bluetooth interface. The fact they have removed the physical connector is stupid, but it absolutely does not mean you need to buy headsets from Apple.
>making the battery not removable ... means the new phone owners will HAVE to buy a whole new phone every few years
This bollocks again. It may not be something you can do on the train on the way home from work (which may be a desirable feature), but it is trivially easy to replace a battery yourself, or pay someone in pretty much every shopping centre or high street to do it for you.
Do you also throw out your car every time it needs a filter changed? Because that's a much harder job to do.
Same is true of ejector seats, oil rig evacuation systems, parachutes, emergency egress systems from test aircraft... That is kind of the nature of emergencies though. You design and test the best you can, but every emergency event is unique, and will pose unique challenges.
Did you know, for example, that for an uncontained rotor failure on a passenger plane, the manufacturer should calculate the percentage change that the plane will be lost. The recommendation is that there should be a 1 in 20 or lower chance of catastrophic loss. But that's only a recommendation, not a requirement, and is only based on a desk-top analysis. No testing is ever done, and when these things happen in the wild, it frequently is extremely destructive, in ways that nobody really imagined.
So, yes, if things go badly wrong, then things have gone badly wrong.
Haha - long thin things look like willies! ROFL!!
> And it starts with US leadership.
> And you begin to understand that the US leadership is completely F*CKED UP.
That's nice. You do realise you're on a a UK website though, right?
Yes, I agree - the concept of a thing connected to the internet is like the concept of a thing with a screen - very useful, lots of applications, a sign of technology advancing. But the buzzword "IoT" means, on the whole, leveraging underutilised potential while moving forward synergistically.
It's not the concept that is wrong, it's the marketing. Products which actually do something useful don't use the label, which means it is only used by useless tat.
Chromecast - works well, doesn't need lots of set-up, works on existing wifi with existing phones, cheap, replaces expensive boxes / services.
IP Webcam - Same as above, replaces multiple devices from security to baby monitor.
Networked storage - Same as above, replaces more complex NAS solutions, allows access across the net via simple setup & app.
I could go on, but that is 3 solid examples of good devices, which can fit into existing networks and provide real solutions, often cheaper and simpler than the alternatives (or the devices they directly replace). All can be properly secured very simply.
Where I agree with you is when the proposal is the opposite - think "smart" lightbulbs which are more expensive, offer no measurable benefits, are harder to use and solve no existing problem. But used properly, technology and networking can and does solve many problems very nicely.
I believe (and have said before) that the problem is in the marketing. Devices which work, fill a good role and are wanted don't need labelled with silly "IOT" style marketing. They're just called what they are - chromecast, NAS storage, etc. So anything which proudly boasts that it is an IOT thing is, probably, going to be a waste of money.
> Scroll down that page and click on the link to their technology partner, Eseye, for an example of the seriousness with which these IoT companies take security
I did that, and landed on a page advertising some sort of internet of things protocol / hardware. Care to elaborate, or am I expected to read through 2 or 3 pages of marketing to get the point myself?
> Over time the battery will degrade after charge/drain cycles. Modern battery chemistry is better than it was years ago, but a decrease in battery performance is observable within the life of the phone. I shouldn't have to replace a perfectly good phone just because a single component is getting towards the end of its natural life
I agree with this, but in reality, it is relatively easy to swap a batter in most phones. A few tiny screws and a bit of sticky tape, and it'll come right out.
For the once-every-3-years operation taking 20 minutes, I'm willing to take the design benefits of not having a "replaceable" battery. As others have said, a power bank is a good enough solution in the situations where a full charge isn't enough.
Weren't the people running the museum in Auschwitz also in the news recently, complaining that pretend digital monsters appearing in nazi gas chambers was a bit... insensitive?
I think the onus should equally be on players to know better than to be using their phones in places like that, but that doesn't excuse the game's manufacturer.
RUD (Rapid Unscheduled Disassembly) is the correct acronym here. Take note Reg. Vague enough to be entirely correct, but most often associated with massive (and very real, supersonic-combustion-of-explosive-material type) explosions.
That Flickr case is totally different. Germany ruled that they should do XYZ in Germany, so they did. They may be a US company, but they were operating in Germany, and complied with German law in their German operations.
I suppose the equivalent would be a US court ruling that this ripping site should be blocked in the US, which would be perfectly reasonable. But attempting to actually alter the behaviour of the site, the company or their operations outside the US, through the use of the domestic US court system, is meaningless.
>I don't have a problem with certified products having chips, it's locking out third party products that do not have the chip that I would find troubling.
Oh, but they won't do that. Not initially, anyway - that would lead to too many returns. They'll just display a message, which you need to dismiss before using the machine (every single time) that there are non-
genuinecertified materials being used. Of course, this will also invalidate the warranty, and the fact it's happened will be stored on non-volatile memory.
A few years down the line, of course, there is a chance a firmware update might change that behaviour, but only after the items are out of warranty anyway. So, on 28 September 2019, you might suddenly find the printer stops working for all consumables without that chip. HP will of course look into this, while locking down any threads requesting answers and failing to respond to any enquiries.
But that's not DRM, of course. That's to enhance the user experience, ensure the product's optimal functioning, and the safety of the equipment. Nothing to do with the current 2D printer situation.
"I'm sorry, the cartridge you have inserted is not certified by HP for use in your machine."
In what practical way is this different from the current situation?
>...doing the deed quickly...
The same goes for leaving a company voluntarily. My last place held me to a 12 week notice period. Not wanting to burn bridges, I instead picked up 3 months more pay for, basically, wasting their time and my own. Very frustrating, and not a good way to encourage people leaving to be up-beat about the company.
Dog whistle is meant to be more subtle. Like, for example, a US
idiot Republican stating "Barak Obama isn't like us," when he means "He's black."
This is just straight out mockery. Much simpler stuff.
(it's ok, it's free, I checked).
I wonder about people like that, whether they are chancing their arm or are actually deluded.
Scamming is sort of understandable... If detestable.
I fear she's serious, and that's sad. I wonder, therefore, about her mental ability and health.
She? Her? Oh, sorry, I thought you were talking about a certain presidential candidate with a floppy wig.
That's the odd thing about Americans, isn't it? Not only do they have a large number of deluded idiots, but an even larger number of presumably sane people who look up to them.
OK, that's interesting, but again are there any examples of two bots battling it out? I'd be interested to know why two bots would both be approved, when the output of one counts as a valid correction for the other to make, and especially so when this forms a circle.
I've scanned through the paper, and can't find one. Are these simple grammar type edits, for example UseTheOxfordCommaBot versus LastListItemHasNoCommaBot? Or is there something more complex? I'd love to see some examples of what's being talked about.
> Surely this would have been apparent as a result of testing. Would it not be normal to perform extensive tests on items coming from the sub-contractor to make sure they performed to specification?
Yes, that's what a Conformity issue means. The design is good, the product did not conform to the engineering. Most Aerospace parts are delivered with a Certificate of Conformity, in which the supplier states that the part has been manufactured and tested according to specification. Something in that process has gone wrong here, and you can be sure the supplier will receive a very good kicking because of it.
That last one would have the added benefit of also preventing those pesky users from trying to replace the single heaven-provided battery.
Now for the real question. How much acid can it hold?
I was out fishing the other day and dropped my After Eights. Now I have a Minty Net of Thins.
> there is a Linux connection. Nearly.
Pengun? I missed it.
Forget the technical part of the test. Hand in hand with it is a certification, from the OEM, that the car does not run a different engine emissions programme when on test from what it uses on the road.
All cars need to know if they're being tested, as the bonnet is normally open, the rear wheels are stationary, the air pressures are all wrong (the car's not moving), etc. These would normally cause the car to throw all sorts of errors, and de-rate the engine to limp-home mode. So, you have a test mode, where it allows the engine to run "normally."
Certifying that the engine is behaving normally during the test is an integral part of the test - if you have to lie about that, you're not passing the test. You're failing the test. It just takes longer to find out.
Same here in Northern Ireland, or at least it was at about 7:30 before I left this morning.
> "[The Raspberry Pis are] incredibly powerful but they're hard to use," she told the BBC.
Hard to argue with that, since it comes without any peripherals, connectors or operating system.
However, it's clearly part of the point. An iPhone is easy to use, and can be made to do many of the same things a pi does, but getting a pi up and running, to fulfill a function, will necessarily teach you about many of the fundamentals of computing.
Which is exactly what it is designed for.
> how would they tackle an electrical fire in any other situation? Why didn't they just use the already-existing steps to tackle electrical fires? It's not as if a Tesla car is anything special in that regard.
It's a fairly non-standard electrical fire. First, it's high voltage DC, not AC, which makes it more dangerous. Secondly, there is no way to isolate it (as they would at sub-station fires etc). Third, it's on fire (the electrical source itself, not just stuff around it), and it's got lithium in it. Fourth, there was no danger to anybody by letting it burn (bearing in mind the driver had already died in the crash), so that appears to have been the safest way forward.
Not sure what else they should have done, really. Seems like they acted very sensibly.
I assume milk floats etc run on much lower voltage, and probably just lead-acid batteries, which have been around a lot longer (and are much less dangerous) that lithium batteries.
I also know that the London Fire Brigade have had extensive special training for the 650V hybrid li-ion powered buses churning around their streets, with packs that make Teslas look like matchbox toys. I wouldn't like to be in or near one of them if something hit it hard.
Since it's been stated that the driver had already died before they were on scene, I think they probably did exactly the right thing, tragic and awful though it is.
Cars crash. People die in car crashes.
In this case, a man has, tragically, died in a car crash. That is all we know.
Actually that's not all we know, we also know that Tesla is working their damn hardest to reduce the number of people that die in crashes, and it's only because of that, that every crash in one of their cars gets reported round the world. If they were just another OEM pumping out dino-powered boxes, you'd hear, and think, nothing. How many people do you think have been killed worldwide in accidents by hitting stationary objects in their perfectly normal cars today?
And yet you want to have a go at bashing them. Well, good for you.
> You can't pack enough brains into a small drone for it to operate autonomously.
Why not? My £40 Pixhawk flight controller can fly a drone autonomously using locations and altitudes from Google Maps. The newer Phantoms have obstacle avoidance built in. Both, and many cheaper options too, can operate a safe "return to base" mode if they detect anything wrong.
It's not like any advance AI is needed. Simply "If something odd happens (eg unexpected obstacle, bad weather, unexpected command...), come home. If something very odd happens (loss of location data, loss of full control), land on a clear space identified by on-board cameras. If something very very odd happens (eg multiple motor failure, or power source failure), turn off motors and, if we're going to be really clever, open a parachute.
> Barrister Robert Tibbo, with minutes to spare before the US Government and media arrived, contacted some of his 70 Hong Kong refugee clients to have Snowden hidden.
What, you mean the US hadn't already tapped his phone? Poor show. So much for a world-wide spying network...!
Oh look, downvotes for correct, if surprising, fact, because someone said something fancy with big words.
There is not much lithium in a li-ion battery, and the small amount there is already on fire anyway. Water won't make it worse.
Here is the official advice from the FAA. The logic, and it is sound, is that you can do nothing about the burning cell(s), but you can cool the surrounding ones, prevent the fire spreading, and extinguish any surrounding material which has caught fire.
(1) Utilize a Halon, Halon replacement or water extinguisher to extinguish the fire and prevent its spread to additional flammable materials.
(2) After extinguishing the fire, douse the device with water or other non-alcoholic liquids to cool the device and prevent additional battery cells from reaching thermal runaway.
The recommended action is to pour water (or other water-based liquid - fizzy or not) over it to cool it down and stop the fire spreading. Or better, to drop it in a bucket of water. If that is done, or a water-based extinguisher used, it would be fairly easy to contain the relatively small fire from a phone battery.
I suspect it's not really possible to properly put out the fire before the energy is all released from the battery, which happens in a matter of seconds once it properly takes hold, so containing it is almost definitely the best course of action.
I heard he already acted Bond once, but the post-production studio wasn't up to the standard he demanded, so he wouldn't allow it to be released.
Aren't self-driving cars meant to replicate the behaviour of humans, most of whom will readily move out of the way of emergency vehicles? So, in what way is this novel enough to warrant a patent?
Or are we going to see patents for every single aspect of the behaviour of a driver, driving a car, when the driver is replaced by a computer?
For example, if the car is going to turn left, it should be able to let people around it know, for example by using flashing lights at the left side of the front and rear of the car. There's a patent, right there. And there's probably one for not driving over children, even if they shouldn't be there, too.
Is it normal for first stages and fairings to fall on cities? Or is this a usage of the word "city" that I'm not familiar with? I'd have thought dropping rockets into densely packed civilian areas would be frowned upon, even in China.
Yeah, this new Apple Samsung is a rubbish phone, isn't it!
Thanks Dan, & AC, that makes sense. I wasn't looking at the bigger picture of the company's history, so was wondering about just this one test. It's possible to do this (test a new finger-prick product) ethically even if it gives dubious results, but that involves duplicate testing with the current best practice, and no reliance on the new equipment for diagnosis. So the harm isn't the finger-price machine, but the lack of any ethics surrounding its use.
While I agree with what you've written, can you explain how a finger-prick test could endanger the recipient? I'm struggling to see why the FDA have a problem here, or what the ethical dilemma is.
Why? Professional conferences do tend to be rather expensive. They all are. It's an expensive thing to organise.
Even if you are skeptical of the subject matter (as I am - see my comment above), that price, for a 3 day conference, is not particularly over the top.
It's a strange thing. I genuinely don't like the concept of IoT - most of what it seems to represent ("smart" light bulbs, self aware kettles etc) seem to represent solutions where there is no problem, at great expense and sacrifice of data/security/privacy.
However... I like my Chromecasts. I like the fact my car can get live traffic updates through my phone. I like being able to check the webcam in the baby's room from the garage, so that I don't have to stay in the house all evening. I like having several TB of storage on our network, which I can access from anywhere. I also like how my phone has become a massively connected communication thingimy.
So I guess the issue, to me, is that IoT has become such a toxic "brand", associated with such nonsense, that I no longer apply it, in my head, to the things I actually find useful. I simply refer to those by name, or maybe something like "internet connected gadgets."
It's surely time serious developers stopped banging on about colour-changing lightbulbs (which cost £100 each, and stop working when the parent company stops making enough profit from the exact model you have) and start actually talking about specific, useful applications to our every day lives. Google seem to be very good at this - when did they last refer to any of their products as IoT devices?
The point is to make a point, that they were in breach.
The point of appealing was to try and reverse that point.
The fact that the actual amount of money is basically meaningless proves that this is about making quite important points, not about the money.
Talk talk may be pretty awful, but their behaviour in appealing this is perfectly logical.