Print

Please login or register.
1 Hour 1 Day 1 Week 1 Month Forever
Login with username, password and session length

[feelotraveller]

Ok, global meta-standard needed: "All sets of standards shall have an even number of standards within them".

(Is that the first and only meta-standard (uh oh!), or is it now one of 2n meta-standards or not?)

2⁰ is even, right?

Or alternatively the meta-standard is the zero point.  Zero being divisible by 2 with no remainder...

[Starver]

Sort of tech engineering, etc.

BBC Radio 4 has a regular¹ Saturday programme called Profile, in which fifteen minutes are dedicated to looking at a given person in the news. Or who ought to be in the news. Sometimes they are trying rather obscure, but Ivve only ever seen the listing announce that the subject is "<Name>", even when the actual intro to the programme starts off wirh something like "Not that many people outside the field of Chicken Sexing have even heard of Obscurey McObscureName, the reclusive Chief Financial Officer of Tibet's third largest poultry farm..." and then proceeds to introduce the man who it appears is the true power behind the throne in countries on seven different continents, including Antarctica, plus the far side of the moon...

(Lest my hyperbolae confuse you, this is a serious programme, not one of those spoof ones like Twenty Players. But I couldn't help myself.)

So colour me surprised to see that tonight the subject is "Lee Kun-Hee (Chairman, Samsung)".   Has it come to this? Is the BBC actually dumbing down?

Anyway, let's see.  It'll be interesting to hear the details of the long-time mogul (and third son of founder) of Samsung, and hear compressed into a quarter hour his long life and current status as, in illness and probably represented by his son, as regent, his company suffers a technical and financial malaise of its own.

(18:15 BST tonight, Radio 4, plus repeats plus iPlayer plus BBC website plus probably podcast in near-perpetuity and everywhere this side of the Great Firewall Of China, in case anybody else is now intruiged enough to listen.)


¹ And frequent, as in every Saturday. Just to cover my internal lexical nazi's usual complaint...

[Reelya]

https://it.slashdot.org/story/16/10/19/1549212/traditional-keyboard-sounds-can-be-decoded-by-listening-over-a-voip-connection-researchers-say

Here's one for you: some researchers have worked out how to decode the audible sound of you typing, to tell what you are typing, without any sort of access to your machine. Though it's been tried in the past, this current method correctly guesses 91.7% of keystrokes. With that, it wouldn't be hard to infer much of the rest, or sample you typing in a password more than once to narrow-down the keystrokes. Any audio/video already recorded with keystroke sounds is thus a potential security risk. The only upside to this tech is that you could steal data from call centers when they call you

[Max™]

MY CLACKITY CLACK KEYBOARD HAS BETRAYED ME?

[Frumple]

That just means the next wave in clacky keyboards are going to be amazing. They're going to be built so the press noise is randomized :3

[Starver]

Reminds me of a scene in Due South ('90s crime/drama-comedy featuring ultra-capable Canadian Mountie being a Flying-Fish-out-of-water whilst serving in Chicago PD for Plot Reasons) where Constable Fraser is at his desk and calls over to pseudo-partner Detective Vecchio(?) to congratulate him on his use of punctuation, or similar, purely from the sound of his laboured efforts on a mechanical typewriter.

(Probably a clip of that, somewhere, to correct my vague memories. But the idea is much like that.)

ETA: also now reminds me of Monty Python's Complete Waste Of Time, with one of the themed programs/utilities was one to play a sound on every keypress of your computer keyboard. There was a "typewriter" sound theme (the same for all keys) but the main one assigned zany "boing"s,"bark"s and "boom"s, etc, to each and every key, individually. Making it 'trivial' to tell which key had just been pressed, because of the (fixed) variety of sounds...

[Reelya]

https://hardware.slashdot.org/story/16/10/18/194231/co2-to-ethanol-in-one-step-with-cheap-catalyst

Here's a happy one though. Oak Ridge National Laboratory have discovered a simple catalyst to turn CO₂ + water directly into ethanol. And it's a government lab, so it won't be locked down with corporate patents.

The team used a catalyst made of carbon, copper and nitrogen and applied voltage to trigger a complicated chemical reaction that essentially reverses the combustion process. With the help of the nanotechnology-based catalyst which contains multiple reaction sites, the solution of carbon dioxide dissolved in water turned into ethanol with a yield of 63 percent. Typically, this type of electrochemical reaction results in a mix of several different products in small amounts. "We're taking carbon dioxide, a waste product of combustion, and we're pushing that combustion reaction backwards with very high selectivity to a useful fuel," Rondinone said. "Ethanol was a surprise -- it's extremely difficult to go straight from carbon dioxide to ethanol with a single catalyst."

Very nice stuff, the catalyst only uses common elements, it uses electricity (which could be from solar) to convert the CO₂ directly into a usable alcohol. This could also compete against biofuels, which would reduce the pressure on farm lands while also reducing food price inflation. Turning CO₂ from a waste product we can't get rid of, into a valuable commodity we use to make other things, is basically the holy grail.

[Starver]

Turning CO₂ from a waste product we can't get rid of, into a valuable commodity we use to make other things, is basically the holy grail.

At the very least, could be a useful way of implementing energy storage. Create ethanol from spare solar/wind/grid energy and then burn it again when demand outstrips local supply of those other sources. Not sure it'd be as efficient as a reversable hydrogen fuel-cell, but might have advantages (temporary ethanol storage vs temporary hydrogen storage, for example)...

[Calidovi]

Would this be cheaper than an average fuel cell? Maybe not now, but in a few years.

[Max™]

Well, I mean, trees turn CO2 into stuff like wood and leaves, farmers turn it into corn and then ethanol.

Also the next wave of mechanicals are optical switches: http://blog.wooting.nl/flaretech-the-optical-keyboard-switch/

[3man75]

Just an idea.

If human work is automated than we could lose alot of jobs. Said humans may need to go back to school to learn and re-tool for newer jobs. But also if humans can just have these robots work then why work yourself?

My point being that if/me/us become obsolete than why would would it be needed for us to be around? Why would leaders WANT us around when some have expressed distaste for Blacks, Latinos, and Native Americans before? This is obviously american cenric but the same can be applied to Europeans and the Gypsies that are still hated over there.

From the point of my limited economists mindset, we the obsolete will fade. Just a thing I wanted to say earlier in the thread.

https://hardware.slashdot.org/story/16/10/18/194231/co2-to-ethanol-in-one-step-with-cheap-catalyst

Here's a happy one though. Oak Ridge National Laboratory have discovered a simple catalyst to turn CO₂ + water directly into ethanol. And it's a government lab, so it won't be locked down with corporate patents.

The team used a catalyst made of carbon, copper and nitrogen and applied voltage to trigger a complicated chemical reaction that essentially reverses the combustion process. With the help of the nanotechnology-based catalyst which contains multiple reaction sites, the solution of carbon dioxide dissolved in water turned into ethanol with a yield of 63 percent. Typically, this type of electrochemical reaction results in a mix of several different products in small amounts. "We're taking carbon dioxide, a waste product of combustion, and we're pushing that combustion reaction backwards with very high selectivity to a useful fuel," Rondinone said. "Ethanol was a surprise -- it's extremely difficult to go straight from carbon dioxide to ethanol with a single catalyst."

Very nice stuff, the catalyst only uses common elements, it uses electricity (which could be from solar) to convert the CO₂ directly into a usable alcohol. This could also compete against biofuels, which would reduce the pressure on farm lands while also reducing food price inflation. Turning CO₂ from a waste product we can't get rid of, into a valuable commodity we use to make other things, is basically the holy grail.

Praise be to basic renewable resources.

[TheBiggerFish]

O_O

[Calidovi]

how are leaders going to stay in power when they have robots that obsolete them?

Well, I mean, trees turn CO2 into stuff like wood and leaves, farmers turn it into corn and then ethanol.

Yeah, but the systems for reverse combustion. Not the ethanol itself.

[Reelya]

I think the main benefit is not just that you can store energy, but you can store it in a way that sucks CO2 out of the air. Obviously if you're burning it then you can say "yeah but it ends up back in the atmosphere" but that misses the point that a fair amount of it will be processed into ethanol at any one time: there would be a lot in production, transit, storage at any time if this became a common fuel. So a lot of CO2 would be locked up in the process itself. You can make plastics out of ethanol as well, not just burn the stuff. Solar/wind plastics.

https://news.slashdot.org/story/16/10/23/0310200/a-british-supercomputer-can-predict-winter-weather-a-year-in-advance

Here's another nice one. Using improved climate models and supercomputers, as well as decades of data, the British have a program that could have forecast the weather one year in advance with a 62% success rate, going back to 1980 data. With models like that maybe things like in Back To the Future II where they tell you the exact time that the rain will start/stop can become feasible.

Obviously, models like this can be over-trained (i.e. trained to a "just-so" prediction that fails on new data), but there are basic established techniques to avoid and test for that, which I assume they'd follow: e.g. you train your predictive model on some of the data, then you use the remaining data, which the model has never seen before, to check it's predictive power. That's exactly the same in practice as training it on past data, and checking against future data. You're just moving the concept of "future" back a few years.

Just an idea.

If human work is automated than we could lose alot of jobs. Said humans may need to go back to school to learn and re-tool for newer jobs. But also if humans can just have these robots work then why work yourself?

A couple of points come to mind. This scenario firstly requires that humans can go back to school and re-tool for newer jobs faster than the robots can. That's been true before, when robots were only good at a few isolated things: a few "islands" of task-types but as robots get better, those islands grow and join up. Eventually, you're treading water to learn new skills faster than the robots do, and there are cases where the robot isn't as good as a human, but they're "good enough" while being a lot cheaper and faster and more consistent.

The second point is that say you buy a robot then send that robot to work for you. Well, other people will be doing that as well, so your robot will be competing for jobs with other people's robots. Then, people who don't manage their robots in a cost-effective way will lose out, and more streamlined, larger robot-owning corporations will squeeze your home-owned robots out of the market. So you might have a situation where you personally own a robot that can theoretically go out and earn money for you, but all the opportunities are already taken by other wealthier robot-owners who have economies of scale going on. Sort of like how bitcoin mining started on CPUs, went to GPUs, then went to ASICS, then went to super-clusters of ASICs. Someone is still making money off bitcoin, and your shitty PC can still theoretically mine it, but your actual share would currently be like 1 cent per decade. You probably couldn't afford the electricity to run your robot on the paltry amount of income it would earn when competing with giant robot-owning corporate groups.

But just think: we already have a solution for not being able to compete. And it's not robots. Say you want a slice of the action on soda sales. Do you need to go set up a small bottling plant and hope you can out-compete Coca Cola? Well not really. Coca Cola has shares which you can buy, and you get dividends. Similarly, if all labor is robots in the future then economies of scale will kick in, with large robot-owning corporations. But shares would probably still be a thing, so you can get a slice of the pie on robot labor by just buying shares.

[Calidovi]

so capitalism will save us from the robot menace