173 posts • joined 11 Jun 2009
Reminded me of a line from Dr Hannah Fry about this, she described it as less a revolution in intellegence, and more a revolution in the scale of computational statistics, that none of these systems are really intellegent, they are just building ever more powerful statistical models.
I still want 1gbps, I just have no idea what I'd use 95% of it for, and I'm the sort of person who has a gig-e switch under his stairs and cat6 in all the walls, so you'd think i'd be the sort of person who'd know what to do with a ridiculously fast internet connection.
What do people with 1Gig connections actually do with it? As you point out, you can do 4k with about 25mbps. So a family of five could all be watching different 4k streams and you'd still comfortably have 800mbps to spare. Figures I've read suggest that you can do 8k video with less than 100mbps each, so that family of five could all stream video that requires a 50 inch TV to even start to appreciate and you'd still only be using a half.
I'd be curious to know what sort of speeds you'd need for a hypothetical future VR telipresence system, but I suspect lag would be far more of a problem than bandwidth for that.
"AI in your fridge" is the new "Internet on your fridge"
In the same way I don't want a web browser on my fridge, I also don't want an AI in it.
Funnily enough it seems to be the same white goods companies just trying to find the latest gimmick to make their otherwise humdrum products seem exciting. I'd rather just have a faster and more efficient washing machine.
Have to agree with your assessment of why Alexa seems to be winning. We have a couple of Echo Dots, and for what they are they are great. They are just hands-free interfaces to some useful functions (which is great when you are wrangling two small kids at the same time). We mostly use them for simple stuff like playing music/the radio, setting timers, getting a weather forecast etc. And I like to use it for checking spellings or meanings when I'm doing the crossword. None of it is life changing, but they are nice to have around, and at £35 when on sale they won't break the bank for a lot of people.
I don't want them to be proactive, I just want them to do what I ask them to do.
Likewise. Very sad to see how this has all panned out. I didn't back it but was hoping that with all the apparently experienced people involved that it would be good enough to plonk down some cash in exchange for some nostalgia.
Sadly it looks like it's turned out as badly as so many other crowd funded projects, only this one was supposedly run by people who should have known what they were doing.
Re: Testing in Oxford
This actually makes me think Oxbotica stand a chance of coming up with something decent.
Most of the other automonomous car developers are focusing on the US at the moment, with all their grid layouts, generally well laid out roads, and wide freeways, that's like driving on easy mode compared to many towns in the UK built on ancient road layouts with narrow roads and terrible junctions.
If Oxbotica can nail driving in places like Oxford or here in Bath then I'll consider it a solved problem. The day an AI can figure out the junction near me that's a five way offset cross-roads (one of which is single lane), on a steep hill, on a bend, with sod-all visibility for the steepest road, I'll be properly impressed.
If it can then make it along the busy street with cars parked on both sides, working out who's turn it is to get out of the way, without getting bullied into staying put forever, that's the day the machines take over the world.
Ah, the "I don't know about something therefore it can't possibly exist" argument.
So you've never seen a DJI S1000 Octocopter then? - https://www.dji.com/spreading-wings-s1000
That has a camera mount that lets you attach your own gear to it, as a number of other professional drones do. Drones with camera mounts and enough power to lift a DSLR are pretty common and used by professional camera operators.
The one in the image at the top of the story looks like a DJI Phantom Vision with a gimbal mount and something like a GoPro, to allow for separate movement. It's not a DSLR in that case, but the phantom Vision has less power so not unexpected
Not just Broadchurch
She was also in the truly wonderful Attack The Block. A very under-appreciated British sci-fi film. And the equally brilliant Venus alongside Peter O'Tool.
She's a terrific actor who's done a lot of different stuff. If the scripts are good then I fully expect her to do an excellent job. I'm really looking forward to seeing her in action.
And frankly, I'm just glad to finally see the back of Moffat. It's just a shame we only got one series of Bill and Nardol, they were great.
Am I being stupid, or does this paragraph have something missing?
"Thanks to the incredibly long distance for the shot, the sniper's target would have heard the boom of the round being fired approximately a tenth of a second before it hit, with the sound wave reaching him 9.88 seconds after the shot was fired. Had the sniper been 250m closer, his target would not have heard it coming."
The bullet spent 9.7s in the air according to the last part of the article, and the sound took 9.88s to get to the target, so how would the target have heard the boom of the round being fired 0.1s *before* it hit?
Have some of the numbers been flipped around or am I missing something?
We have a HH5, and when it works it's fine. But we've had multiple periods where the 5ghz network is completely useless after a few hours, 2.4ghz continues to work fine. Throughput drops to bang on 5mbps each time until it's turned of and on again. There's only one other 5ghz network in range. The BT forums are full of the same complaint but BT just blame anything other than their firmware.
However the latest firmware seems to be working fine now so hopefully they've fixed it somehow.
Or just pay the ine
From the article it sounds like Apple were simply prepared to pay the fine. If there are fixed level fines then this is almost certainly something that Apple could suck up.
It's a bit like Amazon's decision to not implement Mastercard Securecode or verified by Visa. I believe they pay a "fine" of sorts for not using it but have decided that not completely fucking up their checkout process with a poorly thought out UX car-crash is totally worth the cash.
What an offer....
So the options were:
* Volunteer in Feb and get 30 days statutory
* Wait for the redundancy process to kick in, then get 30 days statutory.
Why would anyone pick the first option? Why go immediately when you can carry on being paid to sit at work and browse job sites and tart up your CV?
Re: Said it before, will say it again
Oh god this.
We have Amazon Prime and while it's a nice to have addon for a service I'd pay for anyway (Prime delivery is really useful for me), I can't imagine paying much for just the streaming service.
The film library is pretty rubbish, with lot so of random holes, and a total absence of good older material.
The music industry, on the other hand, have nailed it with Spotify/Amazon Music etc. I happily pay more per month than I ever used to for music on CD because if I want to listen to it I can pretty much guarentee it will be there. It's very rare for me to not find what I'm looking for there.
Getting started early
This feels like a nice gimmick, but clearly not practical at the moment, and I don't think Amazon think it is either.
But Amazon appear serious about this. They've put some effort into those drones and the infrastructure around them. Basically they seem to be ironing out how such a system works in the expectation that the efficiency, range and carrying capacity of the drone will increase over time with improved battery tech.
It feels like one of those long bets that won't pay off for a good few years, but they will be ready if/when it does.
What are people doing that needs fibre?
Genuine question. I rarely hit the upper limit on my 40mbps line.
FTTP feels like HS2, a flashy and expenisve project that will benefit a relatively small portion of the population, when the same money would make a massive difference to the vast mass of people stuck on crap, slow and congested lines who just want "not crap".
Re: Thursday's explosive anti-Brexit judgment
Indeed. This seems to have been the big problem. It was only ever an advisory referendum.
This whole farce has just been a massive pile of political miscalculations stacked one on top of the other, from Cameron assuming he'd win, to Boris assuming he'd lose but become leader, to assumptions about how the other EU countries would react, to the government's assumption that it would have the power to act on A50.
Despite all that, and me being a (reluctant) remain voter, I do wish this would all get settled one way or the other and we can get on with sorting out what the world is going to look like for the next decade. The uncertainty is screwing everything up.
Personally I think the only way to settle this properly would be a second referendum that's actually binding. If we stay because of legal wranglings and political screwups then the damage could be immense. At least with a second binding vote, whatever the descision, we might get some kind of closure.
Re: What's the advantage to the consumer?
Glad I'm not the only one.
The potential annual savings that people claim are frankly minuscule compared with all my other outgoings, and certainly too small to warrant drastic changes like replacing the fridge or oven, or less drastic ones like preemptively changing the lightbulbs.
The irony of course is that those who would benefit the most from reducing their energy bills are generally those without the cash to do something about it. I'm sure a lot of people on the bread line would like to save £20 a month by upgrading their electrical appliances, but they are a bit stuffed by not having the money to do so. The best most of them can do is to turn things off, and you don't need a smart meter to tell you that.
The only significant benefit I can see to these is not having to scrabble around in the understairs cupboard to take a meter reading, everything else is overblown BS.
I'm confused as to how someone can be that confused
I don't get how someone can set up an AdWords campaign without realising they are paying for adverts.
The kid would have had to enter the details of the adverts he wanted to show before he'd have been charged for anything. These are two totally different things. Everything about the AdWords interface screams "you are placing an advert" not "please pay me to show adverts".
Perhaps I'm being a tad cynical but this sounds more like the 12 year old wanted to advertise his band, it cost way more than he thought it would, so he goes with the convenient excuse that he was "confused". He wouldn't be the first 12 year old to think something sounded like a good plan only for it to get out of hand.
We had Mondex on our ID/library card at York university when I was a student. In theory, at the time, it was an interesting idea, you could load your card at cash machines or pay phones on campus, societies could take payment from them with a little calculator style transfer machine, and you could pay for anything on campus (including the payphones and a few local shops) with it.
But it was so, so, sooooo slow. I don't think anyone ever tried to use Mondex at a busy college bar more than once.
Had it been remotely as fast as contactless credit card payments it could have found a niche somewhere, but petty much everyone fell back to using cash.
Nowadays of course it's been made utterly irrelevant by contactless payments. In a world where we have almost ubiquitous internet connections this kind of digital cash solves a problem that doesn't really exist any more.
I want one, but only because our meters are in an awkward spot
I'd like to never have to dig about in the under-stairs cupboard to take a meter reading, but that's the only real benefit I can see.
In reality most people have limited scope for significantly reducing their electricity consumption unless they pointlessly leave everything turned on, and they clearly are not that bothered if they've not done the obvious.
We've changed all the bulbs we can for LEDs, don't leave kit on when it's not used, and so on. I can't really see what else we can do. It's not as if I can heat the oven more efficiently.
Even standby modes on modern kit is much better than it used to be.
The silly amounts spent on this project could have gone on the solar feed in tarriff for a few more years and had far more effect on our energy supply.
The idea that automatic trains are bad or unreliable isn't the correct conclusion here.
It's that *this* system is (possibly) unreliable. Although 750 is less than two per day on (one of?) the worlds busiest and most complex underground railways, so I'm curious if this actually constitutes a "big number" compared with other failures and other networks.
As a counterpoint, the Singapore underground system is fully automated, and from what I understand is very reliable.
One of the problems with the UK is we tend to look at a failing system (education being the classic) and assume that all the problems are inherent to the system, rather than down to a poor implementation, and so we chuck it all away and bring in something new. Because we all know that the best way to fix something is to completely change it every few years...
Also, lets put the numbers in context of the population
Using the 2011 UK census, assuming we only count people aged 20-29 then the number of 24000 cases is about one in 350 young people. It's actually fewer because the population has almost certainly grown since then.
If we include those aged 15-19 then it's less than 1 in 500 young people who will be affected.
Not claiming this is a good number, but as any listener of Radio 4's More or Less should know, the first question you ask is "is that actually a big number?".
For context, how much has the rate increased for those over 30?
Quoting statistics our of context is pretty much the main way to lie with them.
Without a comparison figure these increases tell us absolutely nothing other than fraud is increasing. I would be surprised if the rate of older people being defrauded over the phone or via email wasn't increasing at a similar rate.
I say this as someone who is well over the age range and is therefore quite happy to feel superior if it's warranted.
Re: Very little to tempt
I think you're bang on about much of the tech just not being anywhere near good enough to be truly useful. I'd get a robot vacuum cleaner as soon as the damn thing can get over the various thresholds in our house and can tidy away my toddler's toys. Until then the amount of work that I'd need to do to make it work would wipe out most of the benefit.
A robot lawn mower would be nice if we had a much, much bigger lawn, but frankly the 30 minutes every few weeks it takes to do ours is not worth a grand plus all the hassle of an automated one.
No. Shit. Sherlock.
Most IoT things are solutions looking for a problem, and most seem to make life more not less convenient.
I mean, where's the inconvenience of turning on a light, or using a small piece of metal to open your front door? Perhaps if your house is absolutely massive or you regularly forget to lock your door when leaving the house?
I have a Hive, which is great, but only because it's the first thermostat I've had that I can re-program without having to get the damn manual out every time, and being able to put it in holiday mode remotely after forgetting to do it before you leave. Other than that it pretty much just works as a thermostat, it doesn't try to be too clever, which I like.
I'd like to see a more detailed comparison
One of the differences between CS and the pure sciences is that if you want to put on a science or hard engineering course you will need expensive labs, associated technicians, and supplies. Any institution with a computer room can run a CS course.
This shows. I recently interviewed a grad who had done a CS-lite course at one of the low tier universities, and the poor kid could barely code, with huge gaps in his knowledge. I really felt sorry for him, as far as I'm concerned he was ripped off. I could code better before I went to university than he could after three years.
I would like to see a comparison of employability of CS grads taking into account the institutions they come from. I suspect we might see the difference narrow when you just include the better institutions.
The comparison with the other STEM subjects is also instructive regarding what CS courses don't teach, which is the practical craft of programming. If you do a science or hard engineering degree you will be taught how to conduct experiments, record your results, keeps notes, how to use the lab equipment, how to actually do the practical side of your subject as well as the theory. These things are taught all through your course. When you graduate you would at least expect to have enough skills to get you started in a commercial lab.
The equivalent skills in CS, like how to use source/version control, test your code, work in team environments, document your code properly (not an academic project writeup), these seem to be rarely taught in CS courses, and if they are they are done as a one off, when they need to be woven into the entire course (e.g. submitting coursework through the source control system, with complete version history and test set).
Re: a geodesic past the media
The BICEP2 team were just as much to blame. They even staged a video where they went to tell Andrei Linde about their discovery.
Yes, journalists get the bit between their teeth and tend to gloss over a lot of stuff, but the BICEP2 team screwed up badly by over-hyping what they had before it had been checked.
I'm glad to see they are taking their time.
After the fiasco with the BICEP2 team a couple of years ago it's nice to see a science project taking their time to make sure the data is right and the conclusions are sounds.
Fundamental science doesn't need hyperbolic announcements, it needs a measured and careful approach to announcing findings with lots of cross-checking, otherwise they end up looking stupid.
Take the NATS model perhaps?
While NATS is not a perfect example, it's ownership model is interesting as an alternatuive to the NEtwork Rail approach of total government ownership.
While the government owns the golden share (so retains control and preventing a selloff), about 45% of NATS is (last I heard) owned by a group of the major airlines. Effectively its customers are also its owners.
Make OpenReach a company owned (at least to a large degree) by all the major ISPs, break the old link the BT once and for all, make its owners its customers, and OFCOM acting as arms reach regulator.
That way companies like Virgin and Sky along with the government can actually invest in the infrastructure directly in a way that they can't really at the moment.
Re: Not this perennial nonsense again
I wondered about that too.
Different forces have different needs and requirements, the idea that one radio is necessarily right for all of them may not be right (I don't know if it is, but is seems possible).
Also, are the cheap radios actually good value? Perhaps the more expensive ones wer harder waring and less likely to break down and the plod in Yorkshire did a lifetime TCO calculation that meant spending more on the hardware was worth it over the long run.
In a checkin queue at LAX
I was flying back from a trekking holiday in California that day. We were all standing in the checkin queue reading the papers, which were all about how London had got the Olympics. We found out what had happened from an Aussie family behind us in the queue. I think we spent the rest of the time at the airport trying to get in contact with friends in London, but of course texts were taking forever to get through and I didn't get replies from most of them until we landed. One of them had a lucky escape by getting an earlier train than usual.
It felt like the opposite of 9/11, where we watched everything happen in horrific realtime. On 7/7 it was just bits and scraps of information trickling out. Thankfully all my friends were fine, although one would have probably been on one of the tube trains if he hadn't changed his routine that day. So very lucky.
Re: Changing direction...
I assume the idea is to use trajectories that combine the push outwards from the sun with gravitational slingshots from other planets. Although I am nowhere near good enough at maths to know how you might do something like that.
So kind of like combining the wind with currents on the ocean? (I'm not a sailor so that analogy might be bollocks)
Likewise, thanks Tim
Likewise, thanks for a very clear and concise explanation of something that the mainstream press have largely glossed over. I despair of the major media's inability to explain, even in general terms, what's going on with things like QE
I always like a good TW post. Even when I am in disagreement with Tim about some aspect of politics, I still get to learn something.
It's also nice to read stuff by someone who is happy to accept that you can disagree with someone without them been 100% wrong or having no insight worth listening to.
Sadly being a leftie in the UK at the moment can be hard work. Any suggestion that someone's sacred cow might be nonsense, or that some highly simplistic view of a complex problem might be wrong gets you shouted down.
It's all a bit too "People's Front of Judea" over here on the left at the moment :/