Conversation started Aug 22, 2017 at 20:00.
Aug 22, 2017 20:00
Hi and welcome to Mithrandir24601's AMA. His introductory meta post is here. Please keep unrelated discussion out of this room for the duration of the AMA and use the backup room. Take it away, @Mithrandir24601!
phew - back just in time.
@heather So am I! :P
Anyway, Rytsas Tolvis!
rytsas!
Right, so first question: role of religion/theology
Or, perhaps I should introduce myself first?
Aug 22, 2017 20:02
a quick introduction might be good @Mithrandir24601 ::shrugs:: up to you.
I'm a PhD student, studying quantum engineering and my PhD topic is most likely going to be on simulating non-hermitian Hamiltonians on a quantum chip. I've got a Masters in Mathematical Physics and a Bachelors in Physics with Computer Science. I've got a variety of interests that include music, reading, tea and chocolate. I'm also into religion/theology (there were a few questions on this)
More specifically, academically, I'm interested in things like quantum computing and measurements. Currently, I've just finished a really interesting project on looking at noise in what's known as t-designs, which are used in randomised benchmarking of quantum chips
Oh, and I've just came back from the seminar that's on the star-board!
Anyway, first question:
**What role has religion/theology played in your studies?**
Directly, none, although it has influenced things that I'd consider looking at due to the people I get to know being involved in that, such as medical physics
Indirectly, possibly, as a large number of my friends are from my time as chapel clerk, which obviously influences decisions to some extent, but this is a lot harder to quantify
Does someone else want to introduce the questions?
vzn
vzn
@Mithrandir24601 hi thx for your session! suggest you cover all the questions you want in lull periods but maybe try to encourage participation by audience at this pt?
OK, has anyone got any questions on that?
well, i do have a question, not quite related to the one you were discussing.
what do you think about the recent launch of a quantum communications satellite by a researcher in China?
(i mentioned it earlier in here; i can dig up the reference if you'd like.)
vzn
vzn
@Mithrandir24601 so is your family religious? do you find no conflict between science/physics vs religion, or any thoughts on that?
Aug 22, 2017 20:14
@heather I'll answer @vzn first, as it's more relevant to the current topic
sounds good.
@vzn Yes, although they'd be more typical Northern Irish Presbyterian, while I'd consider myself a lot more (high) anglican
Personally, I don't see any conflict - my Director of studies as well as other students I knew as an undergrad were (are) also religious. On the physics lab of where I did my undergrad, the phrase "The works of the Lord are great; sought out of all them that have pleasure therein" is inscribed on the front entrance
vzn
vzn
please cover your dwave visit at some point! a whole wk? was geordie rose there? did you meet him? any opinion on the aaronson needling of the company? etc
So, I spent my entire undergrad learning from a bunch of religious and talented scientists as well as being friends with people like theologians, the dean of the chapel, the vicar etc. and we also had long (several hours over port, into the early hours of the next morning. The latest I got back from one of these was at 5am) discussions on things like this a few times a term
@AccidentalFourierTransform It is. It's also something that I've really realised I've taken to heart, along with things such as the meaning of 'Logos' (often translated as 'the word', which is the most inadequate translation ever)
Aug 22, 2017 20:22
what translation would be more adequate?
Essentially Logos is better translated as something more along the lines of 'the logical argument', which completely changes things
How would mathematicians view the Bible if one of the chapters started with 'In the beginning, there was logic'?
but anyway, I digress
@Mithrandir24601 that's...actually pretty interesting. i had no idea, thanks for bringing that up!
@heather It's essentially how the Gospel according to John starts
Aug 22, 2017 20:24
@Mithrandir24601 Indeed - "In the beginning, there was the Word, and the Word was with God, and the Word was God..."
or, rather - "In the beginning, there was Logic..."
"... and the logic was God"...
Things like that have definitely influenced my view on things
Anyhow, I should move on - I'll be here a while afterwards, so we can get back to this later if people want
sure.
I'll mention the satellites next
I did an essay about this :)
Aug 22, 2017 20:26
=D
It's AMAZING
reference for those who are interested: Physics Today, vol 70, #8, pg 14-16
Although, technically, all the stuff they did is possible from off-the-shelf equipment and slightly adapted telescopes, with an absolute ton of money :P
The tracking system already existed and has been used to track a plane flying along and transmit quantum signals
was that the research done by Dr. Wong's group at MIT?
Aug 22, 2017 20:30
There are additional effects from the atmosphere, but they don't actually have that much of an effect, surprisingly. Atmospheric attenuation is really low as light's only attenuated in the first 10km or so of the atmosphere
@heather They might have done some of that, I'm not sure, but the paper I'm referring to is actually from nature.com/nphoton/journal/v7/n5/pdf/…
ah, i see, thank you.
@heather So it was really one of those moments like a lesser version of the discovery of special relativity when all the background physics was there, but it took someone really good to bring it together (and in this case, money) to get it working
you said you visited D-Wave; do you agree with Scott Aaronson's comments about them?
So, at the start, what happened is that D-Wave had a bad PR person (or maybe it was the CEO) who overhyped the thing way, way, way too much
and I really agree with Aaronson when he just hammered into how bad this was
@dmckee Can you please join me in the second chat?
Aug 22, 2017 20:39
But a year or two ago, this changed and suddenly, D-Wave is no longer overhyping as much. Since this happened, they've also got a lot closer to where classical computers are (for what they can solve) and so, they're not there yet... But they are close
vzn
vzn
@Mithrandir24601 what was your visit like? when was it? who did you meet? who was there? etc
And I agree with whoever it was that claimed that D-Wave have now brought the fight to the classical computing people. They have 2000 qubits and are nearly comparible
@vzn Who did I meet - I remember the faces, but not the names :P One person who gave a talk was the CFO (although this was brief)
vzn
vzn
@Mithrandir24601 "nearly comparable"... general feeling or do you think there is a particular study or experiment that tends to support that?
couple quick questions (you don't need to answer now, but i have to leave soon):
1. when did you start your PhD studies/how far into them are you?
2. Why don't you like Python?
yeah, perhaps we should move on to the next question?
Aug 22, 2017 20:43
@vzn There was a paper where D-Wave did something faster than a classical computer and it actually took the classical CSs some time to come up with something faster - to me, that was the turning point
vzn
vzn
@Mithrandir24601 yes ok there are a lot of papers trying to do benchmarking, if you can point at one you thought was influential/ turning pt itd be great...
3. which realization of quantum computers are you most interested in (superconducting, optical, etc)?
(although I should stress that D-Wave aren't there yet)
@vzn Benchmarking it is a hard thing. I agree with the people at D-Wave who said that D-Wave+classical is better than classical alone essentially
anyway, thank you for doing the AMA; I'll read the rest of the conversation later with interest =)
Not always faster, but e.g. more accurate
vzn
vzn
Aug 22, 2017 20:45
(ps the buzzword for all this in the news/ science is so-called "quantum supremacy" in the sense of whether a classical or QC is "faster" aka "supreme")
@heather Thanks! I'll answer your questions at some point
So there are graphs where classical alone gets down to the best solution slowly, while D-Wave gets nearly there really quickly, at which point, classical can take over to cut the time down considerably
@Mithrandir24601 What language was that actually written in?
@DanielSank Originally in Greek
(John was apparently trying to write something that would appeal to Greek audiences more than the other gospels)
@Mithrandir24601 I have thoughts on that.
@Mithrandir24601 I see.
I had a college roomate who studied classics and was quite a student of theology.
vzn
vzn
please also talk about your very notable photonic chip work at some pt, maybe QM noise/ weak measurement :) (rats not much time left) :(
Aug 22, 2017 20:49
@DanielSank At this point, you have to look at the context of what this means and how it affects things :P
He made a big deal about people arguing over the meaning translations of translations of the bible. Quite interesting, actually.
@vzn I don't have any notable work :P
@Mithrandir24601 Eh?
@DanielSank Oh, as in, thing like 'how did writing for a Greek audience affect what John was writing?'
vzn
vzn
lol this physics AMA is turning into a religion AMA who would have guessed :o o_O
Aug 22, 2017 20:50
@Mithrandir24601 ah
Oh woops, are we still in the AMA?
apologies
We can get back to that later then :P
::shuts up::
@DanielSank I'm good with it, but there's not a lot of time left :/
Is the IBM's quantum experience worthwhile? Do they really run your code on a real quantum computer and how long does it take to get the results?
@dmckee If you're reading this I would really appreciate some help...
Aug 22, 2017 20:52
^ this question pls
@ThomasKlimpel Yes, yes, for me (being in the UK and hence running things at night in the USA), <1 second
@ThomasKlimpel Although you can also run it on a simulation, which can also be useful
vzn
vzn
@Mithrandir24601 its 5 qbits right now right? with plans to upgrade to ... x? did you write any custom code?
also, at some point I'd like to try and guess your real name :-P
@vzn This is a good point to mention stuff I looked at on t-designs :P
So, yes - I simulated and ran some very basic t-designs on it
vzn
vzn
yep ~5m left on the 1hr, hope everyone can stay later for continued dialog/ fun etc :)
Aug 22, 2017 20:55
At some point, when I get time, I'll run fully blown 3 designs on it and see what happens :)
vzn
vzn
@Mithrandir24601 what is a "t-design"?
I'm actually really, really excited, as calculating gate-set fidelity of 2-designs seems to be intractable classically
@AccidentalFourierTransform How about A. Smart?
@vzn So, you want to test how well you can implement a gate on your shiny new quantum chip
But not just any gate, all the gates
So, how can you do this? Well, 'all the gates' are a continuous set (OK, group) of gates, which, therefore can't be done in a finite time
However, a t-design is a set of gates/circuits that gives the same statistics up to t^th order (more or less) as the continuous group
So, if you want to find the expectation value of running a state through all possible gates, you randomly sample using gates from a 1-design
(not all possible gates at once, but randomly sampling from all possible gates if you had an infinite time)
Aug 22, 2017 21:00
@Mithrandir24601 Is this randomized benchmarking, or something different?
vzn
vzn
@Mithrandir24601 sounds like a big part of future research. ok amazing, thats 1hr already, time flies when youre having fun (or trying to figure out religion...!) everyone is free to stick around, hope Mithrandir can stay, its now open chat. would like to hear from everyone who attended, please let us know any reactions or just say hi :)
@DanielSank It's almost randomised benchmarking. Randomised benchmarking takes a t-design and does the same thing, only implements lots of gates at once (or something like that)
Personally, I view it as the theoretical version of randomised benchmarking, but I may be wrong
The problem with describing a t-design in that way is that it's a mathematical relation
$$\sum_{U\in \mathrm{D}} \left(U^{\dagger}\right)^{\otimes t}\otimes \left(U^{\mathsf{T}}\right)^{\otimes t} = \int_{\mathrm{U}\left(d\right)} \left(V^{\dagger}\right)^{\otimes t}\otimes \left(V^{\mathsf{T}}\right)^{\otimes t} dV$$
Is a t-design
(where D is the t-design)
Probably won't follow this.
Although it gets a bit more complicated when adding noise
At least, not without more discussion of the notation.
vzn
vzn
Aug 22, 2017 21:07
more key info not covered from chat logs
@DanielSank Yeah, it's... 2 pages of explaining things like that to get to the above definition...
vzn
vzn
Aug 11 at 23:10, by Mithrandir24601
@EmilioPisanty Bristol Quantum Engineering CDT
yeah ok let's forgo the details then.
Ok but I understand that you work on randomized-benchmar-ish stuff.
Which is cool.
vzn
vzn
Aug 4 at 23:44, by Mithrandir24601
we will apparently have an 8-waveguide universal chip. This is at least 3 qubits, but apparently, there are things you can do to get more qubits from the same number of photons. Again, I don't know if you can do this on a universal chip... IBM actually has 16... I'm at Bristol :)
vzn
vzn
Aug 22, 2017 21:08
Aug 11 at 23:19, by Mithrandir24601
My primary supervisor will be Anthony Laing. One of his post-docs and apparently someone from America that I've never met before will be secondary supervisors
@DanielSank The basic idea is that U is a unitary, U(d) is the Unitary group in d-dimensions, the integral over that is a badly notated Haar measure
What's $t$?
@vzn Yeah, so I've got more details on this now :) (by the way, 'we' means 'other people at Bristol and not me' :P )
an integer?
@DanielSank yeah - it's a t-design, so for a 2-design, t is 2
vzn
vzn
Aug 22, 2017 21:09
WA and Bristol researchers make quantum computing breakthrough cio.com.au/article/599964/…
ah
So there is some funny way in which a sum over unitaries in the t-design is equal to an integral over all unitaries in $U(d)$.
@DanielSank This is exactly what makes a t-design special :)
Although it scales horribly with t and also badly with dimension
Right so somehow if I want to know about an average over all unitaries, you can get that average by looking only at a discrete subset?
or something
Yes! Exactly!
okey dokey
Aug 22, 2017 21:13
@DanielSank the smallest 2 dimensional 2-design size is 10. The Clifford group [modulo phases...] is a common 3 design and the Paulis are a common 1 design
and a t-design is also a t-1 design
Ah.
Wait, what?
oh
ok
vzn
vzn
@AccidentalFourierTransform can you update the meta post to add this to the full list of talks? youre so good at it :) physics.meta.stackexchange.com/questions/7783/…
vzn
vzn
@DanielSank does Jess Riedel (TBD AMA date) have a SE login?
@vzn I dunno. Ask him.
vzn
vzn
Aug 22, 2017 21:16
oops see his ID linked on post. anyway any news on that one? think maybe just take it off meta post for now, its been around on there so long :|
@vzn lol, I did that the last two times :P
@vzn So, we have an 8-waveguide (3 qubit) universal chip. This will most likely get bigger as time goes on. Bristol also has a 2 qudit chip with some large number of dimensions
facepalm
@vzn What benefit is there to removing it?
vzn
vzn
@ACuriousMind thx so much :) didnt seem like mods were all into this AMA thing sometimes :|
he's a mod, but he's also a user
vzn
vzn
Aug 22, 2017 21:18
@DanielSank its been on there for months, has he even visited chat? this is the "recruitment challenge"...
he protec, but he also attac
3
this chat is a meme
half the people can't into memes though
And finally(?) @heather - I started my PhD studies about a year ago (OK, 11 months). I'll be starting my actual PhD research in about a month (the stuff I've done so far are just small projects - I'm hoping to get a paper out of the t-design stuff at some point though :) )
I should find a message by ACM helping someone, and then another one being harsh to someone
vzn
vzn
@AccidentalFourierTransform sounds like an advertisement for...? :P
Aug 22, 2017 21:20
the latter should be easy to find
I know how to trigger him
Essentially, python is slow and a scripting language, not a programming one
@ACuriousMind Germans like algebra because they're bad at math
vzn
vzn
@AccidentalFourierTransform lately 9/10 written to me would be the latter :P aieee o_O
(though I just don't like it for some reason - just looking at it puts me off it)
Aug 22, 2017 21:21
@Mithrandir24601 what about the name thing? may I guess? its probably something with a "J"
also, have you read Notre-Dame?
@AccidentalFourierTransform Nope and nope
@AccidentalFourierTransform you should do a three panel c meme
(as in, you can guess, but it's not 'J')
something algebraic geometry...stack?
he also mod the stac
vzn
vzn
Aug 22, 2017 21:22
@Mithrandir24601 thx again that was entertaining + great, there was so much left to cover! didnt even get into why you dont like python :P
@Mithrandir24601 well, are you a he or a she?
@heather Optical (continuous variable because it's absurdly promising and other optical stuff is just interesting, although not quite as promising to my mind) and superconducting
@AccidentalFourierTransform He - I'm the guy sitting in the corner drinking tea. I always have tea :)
@vzn Some other time :P
@AccidentalFourierTransform nope
Aug 22, 2017 21:27
@ACuriousMind Is the AMA still on officially?
@AccidentalFourierTransform I'll make this a little easier - it's none of those names
@vzn Thanks! I'm going to make dinner now (It's nearing 11pm...). I'm generally around quite a bit and willing to answer questions :)
 
Conversation ended Aug 22, 2017 at 21:29.