Feeds

* Posts by D.B.

10 posts • joined 1 Apr 2010

Microsoft's Brit kid Cortana lands on UK WinPhone 8.1, but China's is the real cutie

D.B.

Did anyone else notice...

...that in that advert, siri is much more impressive than cortana? She's able to recognise the user's request, recognise cortana's reply, recognise that cortana's reply comes from a second speaker *and* that this speaker is a phone, is aware of her own limitations, and at the very end forms an opinion about what she's heard.

Only Microsoft would try to make their glorified note taking / calendar update software look good by putting it next to a mockup of something behaving like strong AI.

2
2

The rise of the humans: How to outsmart the digital deluge

D.B.

I read this yesterday while sitting around in London Liverpool Street waiting for the trains to start running again (shout out to any other el reg readers who were trying to commute out of there yesterday - took me three hours to get home).

Free feels like the right price to me. Had I paid £7 for the paperback version I'd have felt pretty hard done by. The most interesting points have all been covered in the above review - there really isn't much to it beyond what's already been mentioned. Oh, except that fridges are going to be connected to the internet, apparently.

Perhaps I'm just not part of the right demographic to 'get' it. I've certainly never been accused of being a 'Thought Leader'...

0
0

Is that you, HAL? No, it's NEIL: Google, US Navy pour money into 'associative' AI brain

D.B.

"Though these are all (hopefully) obvious to humans, the fact the computer has come up with these associations on its own illustrates just how good deep-learning systems are getting, and how effective they may become in the future."

Where does it say NEIL is a deep-learning system? The paper indicates it does clustering along with SVM classification (along with some semi-supervised trickery), which as far as I'm aware is most certainly not the same thing. The word 'deep' never even appears in the ICCV paper.

2
0

Forstall ousted from Apple after refusing to apologise for Maps

D.B.
Joke

Re: Guess the fairness of it depends on the situation

If we're going to go the conspiracy route let's go all the way, eh?

Cook knew what state the maps app was in prior to release. He's the guy in charge, and apple pays plenty of attention to these things. If he didn't know, he was being negligent. He could have delayed, but didn't.

Forstall, going by the article linked, seems to have a generally good record in the company (a few missteps aside). He's the most like Jobs was. He's apparently a stickler for detail, so unlikely to have given a thumbs up to an incomplete maps app. He's also probably quite powerful in the internal politics side of things, having responsibility for so many of apple's money making lines.

Cook sees Forstall as a threat. He puts him in charge of the maps app, then either provides insufficient resources or pushes for an early release. It's a fiasco, and gives Cook an excuse to send Forstall packing, then divide up his empire amongst others who will prove less of a long term problem.

The Prince, available on itunes now!

16
1

GCSE, A-level science exams ARE dumbed down - watchdog

D.B.

Can't resist linking the following.

http://www.bbc.co.uk/news/education-10794867

Never mind about the bbc story, check out the graph they include showing proportion of A grades awarded at A-level each year, from the mid 60's to now. It says a lot.

3
0

Top Italian OPERA boffin steps down after faster-than-light mistake

D.B.

Re: Sounds more like ...

Yeah, this article seems to say the same thing:

http://blogs.nature.com/news/2012/03/leaders-of-faster-than-light-neutrino-team-resign.html

2
0

Location-based quantum crypto now possible, boffins say

D.B.

hmm again

Nothing about quantum entanglement proves einstein wrong.

Imagine for a moment that you and I start out standing side by side in the middle of a very long road. A trustworthy third person gives us an envelope each, and tells us that both envelopes contain the same randomly generated message. We then set off walking in opposite directions away from each other along the road. At a pre-arranged time we stop, open our envelopes, and read the message contained within. 'Instantaneously' I know exactly what message you are reading, no matter how far apart we are. But no information has travelled faster than light. In particular, the fact we are both instantly reading the same message does not let me instantly communicate any new information to you that was not originally in that message (e.g. anything interesting I've passed as I walked down the road). We can't even choose what the message in the envelope says - the trustworthy third person always does that, and it's always random.

This scenario corresponds very very roughly to a 'latent variable' model of quantum entanglement. Bell's theorem says that such models cannot explain all the predictions of quantum mechanic, but they're still pretty useful as an approximate way to visualise situations like this.

(Any people who actually study this stuff want to point out the mistakes I've made above? I'm always intersted to hear them :D )

0
0
D.B.

Hmm...

"Not only that, there was quite the speculation that it would happen instantaneously - no matter the distance (thereby circumventing light speed restrictions on data transfer)."

This line says to me you've probably got something wrong. Or the claims were indeed wild.

Are you sure you're not thinking of quantum entanglement of photon pairs, as a method of generating secure keys? If so then we can't circumvent the speed of light - the photons still have to be exchanged. And once exchanged we cannot "flip" them to send a message. Rather they can be used to generate a random shared key, which can then be used to send our real message (again, at less than the speed of light). The attraction is that it is impossible* for a third party to intercept the key when it is being transmitted without our noticing.

*under certain assumptions which may not be true in the real world

0
0

Zuckerberg admits working for man claiming Facebook ownership

D.B.

@Jimbo6

Seems surprisingly real.

http://www.orrick.com/lawyers/Bio.asp?ID=96961

0
0

Google to digitize world+dog with People View

D.B.

Goodness...

...quite the exclusive you've got here. I can't seem to find it anywhere else today ;)

0
0