The rise of the robots.

[UncleIan]

ps. where did acid rain go?

If I remember my water cycle...into the sea?

[CryptoPlankton]

And I'd like to know where did that ozone problem go.

Still there https://wattsupwiththat.com/2015/10/26/ ... one-layer/
ps. where did acid rain go?

Although the "real" scientists seem to believe that the 1987 Montreal Protocol appears to be proving to have been worthwhile:

http://www.sciencemag.org/news/2016/06/ ... emical-ban

Whatever sceptics think, surely it was sensible to replace CFC's (as it was doable) anyway. The best thing about it to my mind though, was the level of international agreement - if only we could achieve that when dealing with other potentially global issues.

And acid rain?

http://ec.europa.eu/eurostat/statistics ... statistics

I guess a 90% reduction in emissions of sulphur oxides since 1990 may have helped take it off the headlines.

There's always a place for a healthy level of cynicism, but until Elon and his cronies take us off to other parts of the universe (have humans really walked on the Moon? ) why take any chances when we perceive potential threats to our only home?

Hippie rant over...

[Slarti]

Very interesting article here from the BBC, regarding the growing use of drones in the retail sector and how their speed, accuracy, and efficiency is replacing the slow, expensive humans that carried out their tasks previously -

Flying drones and robots now patrol distribution warehouses - they've become workhorses of the e-commerce era online that retailers can't do without. It is driving down costs but it is also putting people out of work: what price progress?

It could be a scene from Blade Runner 2049; the flying drone hovers in the warehouse aisle, its spinning rotors filling the cavernous space with a buzzing whine.

It edges close to the packages stacked on the shelf and scans them using onboard optical sensors, before whizzing off to its next assignment.

But this is no sci-fi film, it's a warehouse in the US - one of around 250,000 throughout the country, many gargantuan in size: retail giant Walmart's smallest warehouse, for example, is larger than 17 football fields put together.

And these automated drones are now doing the jobs humans - on foot, or operating fork-lift trucks and mechanical lifts - used to do: and they're doing them more cheaply and more accurately.

http://www.bbc.co.uk/news/business-41737300

Cheers,

Itsallaguess

All I can see that is new about that is the use of drones, as opposed to fork trucks or suspended systems.
Anybody who watched the Inside The Factory series about food production would have noticed that many of those places had automated warehouses.

I remember a school trip round one in Basildon in the 70s where they were proud that no human actually knew where anything was, but requests could be fulfilled in a maximum of 5 minutes (I think it was)
It was an insurance company's document storage facility where they had to hold the documents relating to policies permanently, so removing the space required for humans was a good thing and also made the fire suppression systems less lethal.#

Slarti

[Clitheroekid]

The first robot to gain citizenship - http://wapo.st/2gNoETb?tid=ss_mail&utm_ ... bb923c4a4e

It's well worth watching the Showbotics video clip at the end of the article for some fascinating examples of robots in action.

[DiamondEcho]

The first robot to gain citizenship - http://wapo.st/2gNoETb?tid=ss_mail&utm_ ... bb923c4a4e
It's well worth watching the Showbotics video clip at the end of the article for some fascinating examples of robots in action.

Given how many of the pioneer websites on the web were porn, I wonder how long until such robots have a facility to provide 'personal services'

[robbelg]

Given how many of the pioneer websites on the web were porn, I wonder how long until such robots have a facility to provide 'personal services'

I believe you'll find that's already a thing

The link is to The Guardian so its fairly safe ( it's a video with sound )

https://www.theguardian.com/technology/ ... bots-video


Rob

[Clitheroekid]

I would bet a pound to a penny that the difference was because the lawyers assessed claims as doomed to fail, whereas Case Cruncher said they would succeed.

(From the BBC article)

"The lawyers' results might have improved if only experts in PPI claims as opposed to commercial lawyers generally participated."

That confirms my view. Commercial lawyers apply the law, and are used to dealing with rational people who can make sensible judgments.

"Experts in PPI claims" are mostly not lawyers, and are used to dealing with people who are too dim even to work out whether they needed PPI in the first place.

The actual Case Cruncher web page is rather short on detail, but they say (https://www.case-crunch.com/#challenge):

The main reason for the large winning margin seems to be that the network had a better grasp of the importance of non-legal factors than lawyers.

"Non-legal factors" is a euphemism for saying that the Ombudsman is a well known soft touch, who assumes that all businesses are evil monsters and all consumers are poor little snowflakes who need protection. Consequently, he will often make an award in favour of a consumer where there is little or no legal merit in the claim, just because he feels sorry for them.

If the same case were to go to court it would be thrown out, as judges apply the law rather than "non-legal factors". However, I would surmise that Case Cruncher had evidently been fed with lots of Ombudsman decisions rather than legal judgments (which are far more nuanced and about 100 times as long as the average Ombudsman decision). It will therefore, quite correctly, have worked out that claims which would have failed at court would be allowed by the Ombudsman.

The commercial lawyers would not have dealt with PPI claims or the Ombudsman, and would almost certainly be unaware of this bias.

Predicting what the Ombudsman would say in a PPI claim is child’s play, which is why (1) there are so many parasitical firms willing to work on a `no win no fee’ (and no risk either) basis; and (2) most people dealing with PPI claims are thick as two short planks.

All such claims are basically the same. The presumption is that the claim will be allowed, and identifying the parameters for refusing a claim is a very simple task. It’s about as difficult as predicting the outcome of Manchester City v Morecambe.

Predicting the outcome of a commercial litigation case is infinitely more difficult – if it wasn’t, people wouldn’t spend hundreds of thousands of pounds believing they would win.

It would have been interesting if they had also programmed a `judge’ to apply the law rather than relying on the Ombudsman’s decisions. I suspect the lawyers would have won in that event.

Of course AI is at a relatively early stage of development, and I fully accept that in a few years’ time it may well be able to predict the outcome of a commercial litigation case with some degree of accuracy. But it won’t make any difference – we already predict the outcome for our clients, but even if we predict a loss many clients will believe what they want to believe and press on regardless – and sometimes win, as well.

The only way AI would `win’ is if the clients themselves are robots that act completely rationally. I’ve not the slightest doubt that will eventually happen, but by that time I shall be sitting on a cloud (or in a flaming cauldron as the case may be) arguing the position of the boundary between Heaven and Hell!

[Itsallaguess]

DIY weaponised drones - https://www.youtube.com/watch?v=S0GBeOnxA4M

Joking aside, I'm absolutely amazed that these things are allowed to be manufactured and sold. The potential for large-scale harm, considering some of the payloads the larger drones can carry, must be considerable.

I wonder what sort of wide-scale incident would have to happen for the authorities to realise this?

Cheers.

Itsallaguess

[UncleEbenezer]

DIY weaponised drones - https://www.youtube.com/watch?v=S0GBeOnxA4M

Joking aside, I'm absolutely amazed that these things are allowed to be manufactured and sold. The potential for large-scale harm, considering some of the payloads the larger drones can carry, must be considerable.

I wonder what sort of wide-scale incident would have to happen for the authorities to realise this?

Cheers.

Itsallaguess

My neighbour loves his drones: he makes them himself, and has several weird and wonderful contraptions.

He wouldn't build a bomb, though he could easily if he were that way inclined. I suspect he might do something with a toy attack drone, if he becomes a grandad.

He does have some that are sophisticated eyes-in-the-sky, and might fall foul of potential legislation. A few years back (2014?) when we had big storms, that really came into its own, as he did good service for all of us with a quick check on damage to our roofs when some fallen slates were spotted one morning after a rough night.

[Clitheroekid]

I was amazed to see how far robots have come when I watched this video about `Erica'.

Great to see a robot getting stroppy with a human!

https://www.youtube.com/watch?v=_NTj88EdPtM

It's also worth having a look at the other videos that feature her (sorry, `it'!)

When one considers how far they've developed in just the last couple of years it's quite intriguing to think what they'll be like in, say, 5 years from now. I would guess they would be able to pass the Turing test with ease by then.

[UncleEbenezer]

When one considers how far they've developed in just the last couple of years it's quite intriguing to think what they'll be like in, say, 5 years from now. I would guess they would be able to pass the Turing test with ease by then.

We have a 'bot on an IRC support channel that passed the Turing test more than a decade ago, in that it regularly got taken for a human.

It's still active, but has been made less chatty.

[swill453]

Personally, I don't think a human would have done any better.

Disagree. The pedestrian was already in front of the car well before she became visible in the headlights, so the car was going too fast for the conditions.

Scott.

[Slarti]

Police have released two videos showing the moments leading up to a fatal crash involving a self-driving Uber car in Tempe, Arizona, on Sunday.

Personally, I don't think a human would have done any better. I wonder if Infared cameras or similar on the car would have helped (either with human or automation)

http://www.bbc.co.uk/news/world-us-canada-43497364

I've seen people asking "Why didn't the LIDAR spot her before she was in the camera view?"

Don't know if that is a reasonable question, or not.

Slarti

[Meatyfool]

That is the $64,000 question.

It is almost as if the LIDAR wasn't there.

I would have hoped that the car would be fault tolerant: no LIDAR feed = parked.

Meatyfool..

[Ashfordian]

Personally, I don't think a human would have done any better.

Disagree. The pedestrian was already in front of the car well before she became visible in the headlights, so the car was going too fast for the conditions.

Scott.

Can you elaborate on this as I'm not sure what you mean as your comment confuses?

Are you saying the self-driving car was going the road speed limit and a human driver would not have been doing the same?

[swill453]

Are you saying the self-driving car was going the road speed limit and a human driver would not have been doing the same?

The report says it was doing less than the speed limit. What I'm saying is that it was still too fast, as it couldn't stop within the distance illuminated by its headlights on what looks like an unlit road.

Basic programming error I'd say. And I'd hope a human could do better ie not go so fast.

Scott.

[Ashfordian]

Are you saying the self-driving car was going the road speed limit and a human driver would not have been doing the same?

The report says it was doing less than the speed limit. What I'm saying is that it was still too fast, as it couldn't stop within the distance illuminated by its headlights on what looks like an unlit road.

Basic programming error I'd say. And I'd hope a human could do better ie not go so fast.

Scott.

A self driving car does not need headlights to 'see'. Sensors such as Lidar should be doing this. Headlights are a human requirement. See this link for how a self driving car uses Lidar to navigate in complete darkness http://fortune.com/2016/04/11/ford-self-driving-car-dark/

As pointed out above, it was almost as if the LIDAR was not working/there and if this is the case, it should park itself immediately. This is the programming error if there is one.

Many drivers regularly go too fast in poor visibility (fog, rain, darkness) so I'm not confident human's have a better record in poor visibility conditions.

[swill453]

Ok. For whatever reason, it should have been able to stop. The pedestrian didnt step out in front of the car, it was already in front of the car for a significant distance while the car continued to drive forward.

Scott.

[kiloran]

A British company has unveiled a new robotic surgery system, which is expected to operate on patients for the first time next year.

It goes on to say:

"It takes around 80 hours to teach suturing with manual laparoscopic tools and some surgeons find it impossible to master," said Addenbrooke's Hospital surgeon Mark Slack, a co-founder of CMR Surgical.

"By contrast, it takes half an hour to teach using Versius - this will enable many more surgeons to deliver the benefits of keyhole surgery."

Finally:

They are used mostly for prostate, bladder and gynaecological surgery, although their range of operations is expanding.

The Terminator's been laid off from his job as a ruthless dispenser of death. Looks like he's retrained as a surgeon.

https://www.bbc.co.uk/news/health-45370642

Hmm, robots run on software don't they? I can just imagine mid-operation...... either Windows Blue Screen Of Death or a 5 hour delay while the software updates.

--kiloran

[Itsallaguess]

Humanoid 'rescue robot' learns parkour

Atlas, the robot developed by tech firm Boston Dynamics, has learned the art of parkour.

The humanoid has been taught several skills during its development, including how to run, jump over objects and perform backflips.

The latest development shows the robot leap up on to 40cm (15.7in) high blocks without slowing down.

The company has suggested Atlas may one day be used in search and rescue operations, although critics fear it will be used for other purposes by the military.

https://www.bbc.co.uk/news/av/technology-45841771/humanoid-rescue-robot-learns-parkour

Pretty impressive stuff...

Cheers,

Itsallaguess