How to Give a Great Research Talk

Okay so now it's a pleasure to introduce my colleague Simon PJ you think soda from a shirred his talk as well in the course of the years and is one of the most popular elements in the school he will give the first talk of the school and the last one as well today it will be on how to give a great research talk so Simon is a principal researchers in the programming principle and tools group He's also says his research is very much around functional programming languages so he's worked a lot of his research life

on Haskell and specifically the Glasgow Haskell compiler Simon's also very well-known he's now the chair of the computing at schools Society so he's been involved for a very long time and has been instrumental for a change in the UK school curriculum namely making computer science a mandatory science subject and in schools which is a very Great achievement here here in the UK and finally we've been very proud that Simon was elected as fellow of the Royal Society just this year so please welcome Simon okay good well you will note from your program that firstly a

talk about how to give a talk is a somewhat recursive endeavor and also that the last talk of the Summer School appears to be the same as the first one so perhaps there's an infinite loop in this recursion but actually there isn't it's Just a misprint the last thought Willie is different so one this is kind of about how and why we might want to give them a search talk so I think it's helpful first of all to position it and say what are we doing this what are we doing this for I think research

is about communicating with people it's all you know we we have this kind of picture of people sitting in windowless room staring at screens having great ideas but actually for me it anyway most Research actually happens in dialogue with other people and research talk to an important part of that dialogue there there was sort of big part of our communication and I'd invite you to think if you've probably been to quite a lot of research talks by now how many times have you been to a talk when you didn't just you know you sometimes somebody

i sat through that talk while i sat through six talks like how many times you been to a talk but you didn't Sit through it you thought I'm really glad I went to that talk I learned something I was interested in it I came away a bit excited by it think about those talks and what what did that what was this speaker doing that made you feel excited it probably wasn't just that they were speaking to the precise area that you yourself was working working in that can help but it was probably something else about

what they did or said so I kind of like to Invite you to to think about that and in particular for this talk what I want to try and do is to give some sort of simple actionable ideas that I think might help you you've better talks the good news is you know the general standard is not that high so you can shine without without without too much trouble and so as scarlet says I'm a kind of you know professional researcher I'm sort of sharing the ideas that that I've had and Found useful to myself

I'm not a you know I'm not a professional communicator so you're going to get a completely different picture from the professional coaches you talk to but I think that dialogue also might be quite intriguing to you now research is about communication and giving talks is not about the kind of thing of which about which the right answers so it will be much more fun to me if you instead of sitting there though politely and and Stoic ly listening to me you would interact by asking questions or by making comments from your own experience so but

don't don't just save them to the end if you just sit there like bread puddings I will feel that I have failed so you have to help me feel that I have not failed okay so why young so this is the last piece about some why it's worth paying attention to it was worth paying attention to speaking is that when you're speaking to let's say you know 20 Or 50 or hundred people you are consuming you know 20 minutes worth or or half an hour an hour's worth of people's time and that's the s the

commodity which they don't have anymore of rich or poor we all have exactly the same number of hours in the day so by coming to your talk they're giving you a kind of priceless gift that they will never have to give again which is something of their time so it's quite important that you pay attention to it So the first thing is what is the purpose of your talk if you listen to talks you would often think this was the purpose right it was to sort of impress your audience with how clever you are right

and to show them and sometimes if you listen to a talk you would think this person is trying to tell me everything they know about this topic but because it's only 20 minutes they have compressed everything they know about the topic into a sort of Tight you know highly compressed bundle that nobody could understand and that's really not what giving a talks about I think it's about communicating a single idea right so I'd like you to when imagine giving the audience some intuitive feel for the idea they don't need to have a deep technical understanding

of the idea that you're presenting but it's really good if they can go away with some intuition so one one thing I think is helpful to bear in Mind is imagine that your audience contains you know five or ten percent of people who really know about the topic that you're talking about but there's another 85 or 90 percent who know very little about it they may be a computer scientist they probably are computer scientists but they're not gonna know about your particular area of networking or you know static analysis of pointers in programs but and

you'd like all of those people to take something valuable Away with them rather than to think this talk isn't for me I'm going to switch off and read my email you see the point so we want to kind of what can you give them that even a non specialist in your area will take away and think he's valuable to them and you want to kind of make them feel that they're glad they came so what would make them feel that they're glad they came well something that engages their interest in excitement and so you know

kind of Conveyed that somehow now of course that would be easier if your audience was an easy audience so that the audience you would like is like this they have they've just jumped outta bed in the morning there were full of vim and vigor they have been reading your papers for the last five years and are eager to understand the next piece that you have to present to them and and moreover they've earned aces thoroughly understand the technical background of Course this isn't really what happens right what actually happens is they don't know you from

Adam or from Eve they don't know the technical background and moreover they've just had lunch right so they're in that saman and post lunch phase in which if they're not somehow engaged they're going to slump into a kind of dose so your goal in it is to kind of wake them up and make them feel I really want to listen to this talk I'm going to pay attention and the Other thing is you're very politely sitting you know looking with you I can see your eyes like often when you're in a research talk in a

conference about half the audience you won't see their lies because they're looking at their laptops like they haven't been told by Scarlett not to read their email so they will be like and that's not them their fault it's yours right so it's got to make them drag them out of their laptops by thinking I'm going to find something Exciting here and so I want to say a little bit about what to put into a talk and a little bit about what to what not but rather more actually about what not to put into it okay

we searched so fast through at any any other questions or observations so far so what do you put into a talk well perhaps quite a lot but I would suggest not very much so I was just there two things one say something about what the motivation is what are you trying to Achieve and then secondly focus the whole of your talk on one key idea and you need to know what that is so I want to say a little bit more about both of those things so the first thing is what's your what's your motivation

so in it when you give a talk you have that that sort of two minute bit at the beginning at which every audience will give you some leeway they will pay some attention to you at the beginning so it's quite important to kick off quickly So I find it quite a lot of talks I go to I'm paying attention for the first 30 seconds and then the speaker goes into a long riff that appears to be sort of general fluff about the beginning you know I'm glad I'm here and I came on this plane and

he is the general kind of thing I'm doing I kind of lose lose attention so try to dive in really quickly with something that's going to engage their attention so and what is going to engage their attention well if You can say something about what problem it is you're trying to solve and then give some reason to believe that it might be an interesting and and a problem that's worth solving because in you know in computing there are lots of problems and not all of them are worth solving like novelty is not so much of

interest in its own right saying all that later this week and and and then it's helpful if you can give early on some indication That your talk is going to give a worthwhile step forward not just you know here is you know Mount Everest and I'm going to wander around it but here is you know a reasonable-sized mountain and by the end of the talk I'm going to taking you you know to at least the foothills right so give them some preview of what they might get to if they stay with you so here's a

kind of here's a couple of examples I'm going to stand back and look at these but they Don't have these off by heart right so I'm a PL guy so this is you know Java class files are very big give some big figures so it's sort of sort of some brief numbers and so the idea what's the key idea maybe we can use some kind of compression technology to shrink them so they take less time to send over the network and then and then then but then maybe it could be language aware right if it's

a generic compression technology just gzip you might do so bad but if you Knew that you're compressing Java maybe you could do better so so that's a kind of you know you could think there's something vaguely plausible about that idea you say actually this idea works quite well we can do you know 30 percent better than gzip and I'm going to show you how right so now you've got in a very short time you've got a quick overview of the problem and how you're going to solve it well this was another one how we going

to find synchronization Errors and concurrent programs everybody knows concurrent programs really hard to write lots of you know race hazards when you you know mutate one verbal and another and read it in another and I'm going to show you a type system that is a kind of static analysis that's going to help you find at least some of those errors a particular class of the others and there you might give an example of the cut immediately give an example of the kind of error that you might find do Any very quick summary of what you're

trying to draw you're trying to achieve and so that was about the motivation part right that was the first bit so you've got that kind of two or two or five minutes at the beginning to again their attention yes could you put a video there well it would kind of depress so that could be great right if it was a video don't you know say that you were oh I have such demo over people who do robots right but supposing you a Robotics person and you could say I figured out a way to make a

robot walk better you know and you show a video that says here was it walking before you know and then you know is it walking afterwards like that would be that would be highly motivating right so but it's um it's um has to be you know relevant to to your talk like if it's just hey I'm showing videos not so good so if you're a sort of person do video as well absolutely go for it But of course remember typically I'm thinking often of a sort of 20-minute slot in a conference right or you know

or maybe half hour I sort of it's a bit rare to get you know full and how I talk so videos have to be quite short to pay their way yeah anybody else yeah yes please have comments are good oh can you sort of make sure everyone can hear you well don't you speak to me yeah at IBM so it's how they train you and they make you do presentations to all sorts of People so I think one thing that I learned over the years was that any talk or any presentation has to be dynamic

I mean you can sometimes you can prepare for hours or days but when you arrive somewhere and you gauged the audience and the kind of people they are there maybe what you thought was going to be motivational is totally mundane and boring for them so I think you have to also be quite receptive on the go on what they're absorbing and what not and Willing to change yeah you know as you go along the presentation I mean it could be a video or a joke or sometimes it could sound lame so it depends on the

audience I think it's quite a two-way thing certainly being responsive to notice particularly if it's a smallish audience that you can engage with this is kind of really helpful and you should if you know you need to in the end have the confidence to kind of give up a bit on your planned presentation and divert If necessary but yeah anyone else is it good yeah it is really important in your life we see two different people and now you are talking and speaking louder and moving around and using your body to mimic things it's much

easier to follow you up and seeing the motion and you getting out of you and you know we're seeing that that you are really motivated and like so I'm quite a bit more to say about enthusiasm and presentation towards the end of this Talk so yes how you present I'm so the moment one what to present and we're going to get to how to it's super important super important thank you okay let's do a bit more about to what so motivation and key idea so the key idea is really quite important right and when you

go when you leave a talk here's an exercise but every talk you go to this week when you walk out say what was the key idea of that talk that I've learned if you can't articulate it the speaker Not you has failed right so I you can you can you can try this out of my colleagues this week it would be quite a laugh so but in when you give a talk try to make sure that the audience could be in no doubt what your key idea is because really a talk is a kind of

it's like a it's like a pill that you're giving you know the audience is swallowing the pill and it's carrying a payload right of the drug that's going to infect their system and you want to Kind of get that idea into their head so they can't think about anything else like it sort of tuned it can't don't forget like so so but you've got a you wanna you have to know what the idea is because if you don't they certainly won't all right so that's why I say yeah ruthlessly prune the material around the key

idea right so you have lots of stuff you know you'll have been when you give talks you'll have been working for months on a particular talk or Particular you know maybe it's a talk about the paper the paper has quite a lot of material in it but somehow there's a the core of the paper there's a key idea that's what you've got to convey right and and everything in your in your you know took your talk you don't have very long has to be pruned around that that that moment and moreover I think it's helpful

if you're explicit about it sometimes you have a key idea but the audience still doesn't Get it right so you want to you want to be absolutely clear you can have slides at the top that say you know this is the key idea you can you know stand on a chair and say this is the key idea at this moment if you remember nothing else from this talk remember this I get across the key idea and making make sure that the audience know what is that make sense I'm super super important now another thing that

you will often see in talks Is that they kind of try to cover too much right so I call this sort of shallow and broad rather than narrow and deep like so in order to position your talk you have to say something general about the area that you're working with but sometimes people do a sort of you know a horizontal you know a shallow trench that covers a lot of ground but not in any great depth and that's very unsatisfying right what's much better I think is to do a short overview that Positions what you're

doing and then do a sort of deep dive I call it you know producing the technical meat right imagine your audience is sort of hyenas you know they need something a they need a red meat we've got to give them some something substantial and that's this sort of deep bit now that means that there are other deep bits that you've dug like and I put a lot of attention to that you're not going to tell them that's okay Like they can get that from the paper but you have to give them I think part of

what engages a technical audience is interest and enthusiasm and you know we are talking about technical audiences is some actual technical content right and it's difficult to do that because in your in you know computer science is very specialized so to take your audience in 20 minutes from zero to some actual content in your space you'll have to be pretty precise about what you show Them and what you don't why you cannot show them everything does that make sense yeah yeah Bob Oh cost and when you're speaking to other people entirely it's even harder right

so like if you're this happens I sit in this room a lot but and Microsoft interview talks and put people at the interview for permanent posts they're asked to speak to the whole lab so I'm listening to people in you know who doing machine learning I'm a PL guy and I'm struggling To understand them but that but but nevertheless you know it ought to be possible for a machine learning person to explain an idea to a PR guy right even an idea in their in their little patch so it's harder it's harder the more your

audience knows about your area the easier it is yeah okay but I do encourage you don't don't just do the you know we got the Wiswall system and the you know that sort of issue the shallow shallow And but in order to do this but how are you going to present this this techno material well do it by example right so examples are your main weapon in presenting ideas clearly never in my humble suggestion least there's a starting point never present you know the theorem or the generalization first instead produce an example remember we're talking

about some motivation I put up what did I put up a couple of examples right to Illustrate what I meant by motivation and when we're talking about synchronization errors I said well perhaps even at that moment when you're saying what to frame your problem rather than talking about synchronization errors in general you could produce a particular example of a synchronization error and put it by putting right up there so people get there get a hold of it right so everybody in their learning always moves from the specific to the General and yet extraordinarily when we

present to other people we often try to move from the general to the specific it's completely bonkers or the way we all learn is this way right and so that's the way should take your audiences it's not you know a universal truth but it's a pretty good starting point and all that the good thing about examples is they can often when there's a complicated technical point right so sometimes you you do get into technical Something say why does this equation at this side condition you can say it's because this example demonstrates why it's necessary right

you can show an example that says oh now I see that you know the simple rule wouldn't work for that yeah because I mean most of the time people seem to switch off unless it's like a really technical conference so this is let's say Oh what are we gonna what to leave out there I'm gonna skip to a bit about what to leave out so One one thing to leave out is too much technical stuff right so this kind of slide alright and then you know you people put up these kind of slides you could

see the audience going oh they sort of you know reel back in their seats right because and the trouble is that it makes them it's a bit daunting and unless they're really into this stuff they're not reading this I mean if they just makes them feel stupid right but they think I don't understand this The other people in the audience very clever chaps you know they they all understand this but I don't I'm a mere worm I will so you absolutely like and yet right I do think you want to get to some technical detail

right so I find it deeply unsatisfying if a talk just says well you're not really clever enough to understand that you know anything important so I'm just going to give you some generalities so what I would do With something like this is to pick you know one of these rules say you know blow it up big and say I'm just going to show you a little snapshot of a type of a little corner of this system and how we think about it as it illustrates an interesting point and then you may say you could have

this as a backup slide or you just flash it past do you say the interesting thing about this slide is that's the whole type system you know it's in the paper you can read it but I Think it's quite exciting that it fits on one slide like then you're not then that's a whole different thing do you see what I mean so it's a balance definitely no overwhelm them with this kind of stuff but don't go to the other extreme of no no technical stuff at all another example yes but but in this case the

example might be so here these are typing judgments right so I might pick one of them as an exemplary typing judge with but then I might have some little Programs that illustrate its operation yeah so it's all Turtles all the way down through now yeah okay let's go go back how fast can this thing go back not very fast examples yes I'm gonna skip this example because that was just another example of an example let's see this is this is good what did leave out this is even more important so what so what are we

going to put in quick reminder what are we going to put in number one motivation Number two the key idea very good right remember that that's all you have to put in and here's some things not to put in this how often do you see this at the beginning of the talk right early on it so it seems that good scholarship right speaker says thank you for we are very pleased to be here you know it's so exciting to me how come a long way so etc only put here's the drought line of my talk

and then they they sort of say well you know first I'm going to tell You about the flu goal system and then it's shortcomings and then I'm going to tell you about synthetic epi morphisms and at this stage you have no clue what flugel is or what of synthetic epi morphism is and but because the speaker is some you know they're actually quite interested this stuff they can't resist adding a sentence or two about each of these points like all of which is complete double dutch to the audience of that stage right so it's like

you're Taking this precious two minutes at the beginning of your talk and you're just talking nonsense to them right and it's it's it's unimaginably dire and people do it a lot not University I would say you know maybe only about a third of talks do this but but don't yes polished material was so theoretical how do you encourage how do you encourage the presence of pedagogical material that is focused on exactly how do you encourage the how do You edit Gaja , to oppose to focus on examples in a technical field programming language is very

theory based but on the other hand so so here you are so think about it this way you'll hear giving a talk because you have witness a per maybe or being invited to give a talk about something that excites you like somehow um your PR person why so there's something about it that you think is interesting right what is It alright so somewhere in your heart you're beating hard it says this is you know someone which is so important to me I'm devoting years of my life to studying this stuff all you have to do

is to rip it out of your heart and show it to audience right and it's I'm sure it isn't that it's you know it's highly technical Wow that's why I so that's why that's what I'm so excited about it there's something something else right so there the technical stuff is a is a Kind of support to to or something you know for me it's for I'm a I'm a PR guy programming languages guide to I'm interested in functional programming and I think you know functional programming is just a sort of radical and elegant attack on

the entire enterprise of programming that I you know intend over the course of my could upend the whole business of programming well it may not work out like that but at least it will be fun and and then of course you know There's plenty of teton for that if you're building a rocket right there's plenty of detailed technical stuff about how you make the rocket and how you make the spacecraft get into orbit around Jupiter which it has right but then finally the point is it was excited they've got you get the idea I don't

know if there's no easy answer but but go back to go back to basics I would say but don't do outlines no more outlines it's okay to an to do a Bit of orientation slide so you know I notice I have been putting up slides like this that say that tell the audience that I come to the end of a little bit right so actually you know this could be a good place if you've been you know wanting to ask a question and not sure when then you know this could be a good moment it's

a sort of signposts sometimes people are put up content slides like this they don't make much of it they just sort of go by Quickly and they say this is roughly where I am right so just as a quick brief orientation about whether where people aren't at all but don't do it at the beginning and don't hesitate very long on it does that make sense have you you been to talks that do this and have you thought wow I'm gonna stop reading my email and look at this content slide you're not okay here's another thing

oh yes go okay so tell your supervisor that Simon says that you shouldn't I may not Work but but you could try or the other thing is you do put it not at the beginning put it after the motivation right so you've got them motivated and hooked and then briefly you're gonna say in my talk and make a content slide that has three things on it not eight you know and now I know now I've sort of given you a bit of motivation first I'm going to you know show you the key idea then I'm

gonna say a little bit about how well it works in practice you Know and you on you've taken ten seconds satisfied your supervisor and moved on but there is a recording of this talk yeah that's why you can point to the video sending the link yeah yeah yes yes so should you so reading your slides is not good absolutely I would not go through stars to forbid any overlap right because otherwise you think well you know now that you've gotta kind of its kind of split brain think I'm a tweeting this Item and listening so

and you'll get a lot of advice from the coaches about you know how many words to put on slides and so forth and people vary a lot in their styles I tend to be a bit wordy right and and with the other thing to try to avoid if you can is to us a bit or that is to is to treat your slide solely as a prompt for you there can be useful in that way but but yeah at all costs do not just read the slides that is important yeah but you could have Notes

that's fine yes and so this will vary a lot between people some people will be fine with that notes others will want to look at notes others will feel more able to move around dis important things to be relaxed about it oh and but it's much better if you can avoid reading your talk right that part that is it's difficult to do that well I mean so you know if you know your talk really well you can have it completely scripted in front of you but if you can Avoid actually having to read off it

it's very helpful I mean it's it's a this is another personality thing it's quite difficult to just extemporize to begin with but you'll get you'll get used to it but if you're perceived to be fixed on a podium with your eyes on your talk read reading from your talk that isn't just not an engaging way to it you've got to try to avoid I've again yeah okay let's see well what was this when I sit work yes what about related Work so sometimes this is not such an such a frequent sin but it does happen

the people's can spend quite a lot of time talking about related work which seems to be good scholarship like worst of all they can do this early in the talk and that's terrible because it's even more technical than that content slide nobody understands what you're talking about except you know the three you know mister xxb right is very pleased maybe but that's all so but even Putting it near the end uses that sort of precious moments of your talk so rather than having am a related work slide that really is only going to be appreciated

or or learnt from you know how might how many of your audience are going to go away thinking I'm really glad that I you know that there was that slide about related work like most of them going to are going to go away thinking well I'm really glad that person has done their good scholarship And when I read the paper I'm going to find out about the related works in a site all the relevant stuff and really glad that went chat when when somebody asked a question about related work they seem to absolutely know their

fields right and was able to respond in an able and articulate way I'm really glad that when they said yes absolutely we built on the word worm you know the work that Martin Tedeschi did live in Scala which is Pretty amazing you know particularly you know his treatment of you know object-oriented programming or something there so you know take the opportunity to speak well of your as it were competition or more accurately their shoulders are the Giants on whose shoulders you were standing but don't actually proactively force it down your audience yes anything the related

work to the motivation section because some of the related work might have motivated Your current work so saying that this and that happened and so it's a very interesting idea but yeah so good thought but if you were the audience listening to that put yourself in the receiving end was it a direct interest to you who you know who did it and why why what's of direct interest to the audience is the idea sequence so you might say well here's you know here's a problem our Java class files you know a standard you Know you

might think you could compress them with gzip what everybody's heard of GT you might not then say you know and the authors of gzip were so and so biotin it was done in 1976 you might say you know because the idea sequences there's some sort of generic compression technology and then but but but the idea that is talk is that maybe we could use language or where compression technology you see what I mean and then in the paper you cite chapter and verse for the For the competition if somebody asked you about them you could

do but the point the point is so you're trying to focus around the points that the audience is directly engaged with interested in so and sometimes just possibly they will be right it may be that something is so blindingly obviously it'll be absolutely in their minds you know the standard tools that they use you've got to mention okay let's see we talked about technical Detail so we're going to omit it about one thing I forgot to say is one thing that makes it difficult to omit technical detail is that it's you invested so much in

it yourself right drenched in your blood right that that figure right you've been through 30 versions of that figure and you can hardly bear not to show it to audience and yet you have to put it in the paper so it is difficult and I think that probably you know the reason that that Speakers sometimes overdo technical detail is because it's you know it was it's hard one to them and they want to share share something about it but then again think what is it that I'm sharing it's not really the technical detail itself it's

the insight that kind of led me there see if we can extract the insight in some way okay so far so good so that's all about content any anything anybody got any other observations or questions about content yeah a Conclusion you mean you know what you what you've learned from that result yeah I suppose what would you do if you were with a friend in a you know in your office with the whiteboard would you show his up without result without conclusion or will you show result a little conclusion then you look for the kind

of thing you're thinking of what would you do you probably interleave them wouldn't you so I would it's actually a good model what would you do If you were just explaining it to a friend with a bit of paper on a whiteboard trying to avoid thinking of your talk as a sort of formal entity and more as a kind of informal dialogue you know ideally you get into that's true so so yes if you you don't show too many things but I think the force of your suggestion was you might have a key idea but

then you might say well this one of the reasons it's a good idea is it because it applies in situation X and Situation Y in situations edge Y so here's my you know I applied in a situation X and this is what happened that so that's a that would all be a sort of outworking of the key idea as it were yeah but yeah you want to make sure they're still they're still anchored to the key idea okay well let's see well this is being mentioned a couple of times but by far the most important

thing about Presenting a talk is to sound as if that you are interested in it right that is to be enthusiastic about it and it's um it's a you know because if you don't appear to be excited by your talk why will your audience be alright so this is really it's varies a lot between individuals I'm a bit of a sort of exhibitionist so it's not it's not hard for me I know that some people find it much more difficult to know to look their audience in the eyes but Nevertheless if you if you stand

behind a podium and um and you say well I'm gonna give this thought today it's not very interesting you probably won't really want to hear it mean you might want to read your email but but I'm gonna give it anyway so slide one then you know and this is not what anybody actually says but somehow there that you know subliminally you get that message and that's you've got to somehow just struggle with yourself to break out of Out of that feeling everybody feels like a worm right that nobody will be interested in your talk you've

got to set a break out of that and just share the sort of you know the passion in your heart for what you're doing and and with your audience yes everybody back no no you're looking a bit yes you can okay keep speaking loudly so he stalks him the lowest of stones and he's like so almost moved and he's not animated at all and but he is he's Somebody that thousands flock in to listen to so I think sometimes I kind of like God you know so so he gets a free pass so I would

so yes there were counterexample so I and clearly you know the things that he says people have learnt that the things he said says are worth coming to regardless right but nevertheless it's a kind of it's like what enthusiasm makes people dramatically more receptive so it's like think of your audience like sunflowers Right they're all you know sunflowers are nice kind of like this the Sun comes out and they all come out like this and they become sort of receptive to the Sun so enthusiasm literally makes audiences do that they they look at you they

smile they become more open and they've become vastly more receptive to the ideas now of course the idea is maybe no good right and it's possible that you know the ideas are so good that even the most so the droopy audiences will Nevertheless absorb something but you might as well give yourself the best chance was to somebody else yes personally I would strongly recommend moving around conference format is the conference format is that use these there and you sit and you click on your laptop make strenuous efforts to say I'm not going to do that

right so so for a start always bring your own wireless gizmo so that you don't you're not tied to your laptop and certainly that you Know very very occasionally they say we're recording this you know our camera is fixed you must stand here that's very well actually usually they're usually they're not recording you and even if they are that somebody couldn't point the camera so makes join us effort to be allowed to move around yeah while the microphones the other thing you really want a wireless mic if you possibly can yeah yes so I would

I would so it's a judgement you know even without a Microphone I would try the room and see you know can you hear me at the back when I'm speaking without microphone pizza I bet you could probably hear me with that microphone in this room it's fine and then for the scale of talks you're probably going to be doing for the immediate future that will probably be fine but I mean it's up to you right so it but the the key thing is I think you could give an extremely engaging enthusiastic motivating talk standing Completely

fixed in place right it's not this isn't the be-all and end-all it's just that I personally find it easier to move around yeah well yeah it's always about it you know if I was doing handstands you'd need to stop thinking about the talking about something else so it's about it I don't think I walk around a lot I don't maybe people are just being very polite but nobody's told me recently that merely walking around there was one Person I know who's Roger Needham actually who was the lab director and he would do this and that

sort of repeatedly joined the entire talk now that was a little bit distracting I have to say he was a great computer scientist also also stand behind you when you talk to you so yes one reason I like moving around is that it's you know giving a talk with stressful activity and physically moving it helps loosen you up actually so it makes it easier I find Okay um let's see so I said right besides the night before here and I don't quite mean that because I often I did use to do that and actually not

nowadays I found I often write some slides during a previous talk because you know there's a there's a talk and I think it's kind of interesting I wonder if I could modify this talk a little bit and nowadays you can do that the PowerPoint it's great I come from the days when we actually had to handwrite Slides on acetates and had over train overhead projectors but the important thing is that you need to be familiar with them right so it would be it's perfect if you never actually have to look at your slides at all

because you just know what the next slide is right so you the the the antithesis of that is when you spend the entire talk doing this you say oh you know this very interesting rule over here you know I know and you And your point you have your back to the audience the whole time now again that's an extreme version but your goal is to be eyeballing your audience almost all the time just occasionally you may need to say you know this particular point on this particular you know look at look at this particular equation

then you can you can turn them out to try to avoid doing okay another thing that often happens in talks in its almost subliminal is people kind of apologize For something they say you know I was delayed on my travel here my laptop crashed last night or it's probably not a very interesting talk or you know there's something there's something negative about what they're doing sometimes it reflects something that really has happened you know they really didn't get any sleep now it's dawn so they really were sick last night so they really are jet-lagged but

don't share that with your audience right because They while they might be sympathetic they don't think there's nothing they can do to help you at that point right look retentive so it's not it doesn't set you up to be in a good place so just never apologize all right no matter how bad things don't apologize to audit just keep going you're good right because it otherwise it leaves them as an audience member if somebody's apologizing to it you just don't know what to do with yourself right you've Got no way to respond and so that's

no good it's an individual that's quite different right but you know in a group setting it's not so good now of course there are physicals things right so when you're sitting here you know and the previous speaker is you know just winding up their talk and you're sitting there and you're thinking I'm actually not sure that I'm physically going to be able to stand up you know you're kind of breathing fast And you know you can feel a kind of tremor and and you think will my legs actually work apps out there I've really thought

that I thought am I going to be able to physically stand on the stage ah well I simply collapse it's very unsettling but I've seen very very few people collapse on the stage it's quite rare so actually it works out that it you know you can stand so the it is a worry but it usually doesn't it doesn't turn out to be as bad so Here are the things that I do what one is that you know do the sort of usual deep breathing kind of things and nothing that I find helpful is for the

very beginning of the talk that first few sentences right which they're you know the the host has introduced you and like searchlights 150 I swivel and a fixed on you like that moment you're like the rabbit in the headlights that is a good moment to have an exact script for what you're going to say right it Just gets you off to a good start usually once you're going things are okay so when I said about reading a script I would have it I would have that actually memorized to write to know how you're going to

start so you have a strong start that you feel confident about and then then you away yeah I've said the other things oh very yes yes please so I think this is very personal I would I would not recommend starting With a joke I think it's if you're the kind of person who can tell jokes naturally that's great and you know and then go for it but I yes I would be cautious about as it were you know reading up on jokes and thinking I'm going to start with a joke you know it sort of

kind of like if it comes naturally to you fine if it doesn't it's a it's a it's a technical presentation but it's not enough to dinner talk if you like that's it that's Another medium yes the audience laughs and we forgot about everything Costas apologies so I should say I mean humour is good like that if you if you can get your audience to laugh that is fantastic it's just that it's uh it's hard to do by trying if you see what I mean so they it kind of so I think it I don't think

it for me it doesn't come under the heading of a simple actionable idea it's more something you might go Into but yes sometimes people very naturally and you know in a kind of relaxed way can turn a tricky situation into something people laugh at and that's so brilliant when they do that but I I don't know how to turn that into an actionable point yeah altogether right so at that point your slides really are quite helpful right you really can look at your slides and say so to get yourself back on track what am I

trying to say so just try not to save The audience I'd completely forgotten what I was going to say just but but but you can you can have silences right it's okay simply to stop and just have a moment you know a moment's pause while you regroup they're not going to you know that if you stop for you know a whole minute they're going to get a bit restive but simply pausing it's okay and then you can just sort of so yeah don't apologize just pause whichever you are we grew okay we talked about facing

Audience and not pointing I have seen people pointing at their laptop they let it sit on that very much usually point but even pointing that screen is pretty dodgy right because it means that you're forced to put your back to the audience and that means you're you know you're not watching them for questions you're not you're not you're not really engage with them so while you know a laser pointer can be a useful thing I don't use it too much and it's a slightly Dangerous thing too because if you if you are a little anxious

then it turns into sort of lissajous figures you know it amplifies your the the shaking of your hand and so you get little figures of eight on the screen so it's kind of kinda bit not good that way let's see now it's quite good to one you know what to make eye contact with your audience and and it's quite helpful to identify somebody in your audience or maybe one or two people who Know some people instinctively not as the you know as at all it goes long some people just sort of sit there and passively

and doesn't mean they're not paying attention or not interested it's just their their natural body languages but others you know do more sort of instinctive nodding and you know some of you have been doing that like so so you've been asking what's your name I mean yeah that's right Muslims know it's you that's right yes so you've been up Some questions as well like but but you're not a bit and so and not us are quite useful if you can find somebody who's doing that they're people while they're following they all sort of nod but

when they get lost they stop nodding all right so it's great so so you can you can start to kind of say oh so you know is anybody lost here you know because sometimes when materials getting a little technical it's good good idea to do a check and say it would really Help me if somebody asks a question really say Rob how's it going you know did you understand that so if you identify someone in the audience you actually know so and of course by watching the audience you can also keep an eye out for

questions and for me the in conference formal conference formats it's always give the talk questions at the end everything else I'd strongly encourage you to to look for questions during the talk in fact I've got this is What this slide is about so if you can if your audience is willing to engage you in some dialogue during the talk that is a huge victory right if they sit there respectfully until the end and then ask you you know one or two questions it might be okay but it's not nearly so great as when they've made

comments and questions during the talk you know as you have been doing yes go of course you're dealing with antagonistic questions motions that are Designed to show off so this does happen so sometimes yes you get the sort of troll kind of questions who are really kind of derailing your talk so it's it's comparatively rare in my experience right but it's why you might feel anxious about questions and so I think you know stage one is to just you know respond in a constructive way when they pick it start to become persistent you then say

no so there's plenty I could say in response to that question But if I'd have I respond at length we won't get to the rest of the talk and there's you know there's some key ideas I really want to get to can we talk at the end and then if you're really struggling that's when you start eyeballing your um your session chair and you know sort of making signs of distress because they should rescue you right if somebody's being really aggressive but in my my expertise vanishingly we're maybe different They've different things happening different cultures

- you have to just say you know so it's about you've got them in a dialogue you say you clearly know what you're talking about and there's you know there's tons in the paper but you know what button but I'm keen to get to the you know the the key idea that's you know in a couple of slides of time so you can you can just say you you you understand control the other word you want to be responsive but Nevertheless somehow in control so I think it's okay sometimes particularly in a less formal talk

you know in a let's say you're having a seminar in your department - you're visiting somebody else it's about not giving it giving a talk well then you know even truncating your talk not doing some of the material but engaging in a dialogue is way better than saying shutting everything down and saying I'm just going to stick to my script it really Annoys me when people are totally dismissive of questions they sort of blow you off very quickly because you feel they've got their thing they want to say and then really not very interested in

you the audience right they just want to say their thing that's annoying so if as in all things it's a balance but for me I would err on the side of following the questioner rather than blowing off until they you know till it goes too far that's about it yes Right at the back Oh what do you do if 80% of your audience is lost well so again it depends a little bit on the on the circumstances if it was a departmental seminar I'd more or less if I really thought my oldest was getting lost

I'd sort of stop and you know see if I could do sort of whiteboard II kind of things or find out where they were lost because the truth is very little you know the world is not made a better place by you Finishing your talk and having 20 people who just happen understood anything of it it's better just not to finish your talk so I mean they'll come a point when you're listening to talks there'll be a point when you may think well this isn't in my area I was engaged in the beginning and now

it's got too technical and I'm going to disconnect that happens to me a lot particularly when I'm going to interview talks in other areas that's going so for the speaker at that point I Wouldn't want them to think oh I've got a stop and deal with Simon right so because you know that there is a there's a kind of the talk is framed with a particular technical point right and if you can't get to it at all you'd be stuck so but if everybody's lost I'd stop and it's easier to know that in small groups

something I suppose well don't give you much of a chance to repeat because they're quite short typically bit so I would typically do a Repetition like that by way of examples you know here's another example that illustrates the same idea and it's another example that illustrates the same idea so you're not repeating yourself you're just exercising the idea that would be sort of trying to say okay so you know here we are it's a you know it's a half-hour talk and we've had quite a lot of questions and we're you know 22 minutes have gone

but there's there are two things I really want to Show you you know can we just pause and do that and then you but then you then you probably a you know you must kiss me I might need to skip a little bit to get to them right so you've told your audience that you adapting to then write that there are some things you want to to get to because you think it'll be of interest to them but you are you're not just racing through your burger twice the speed yeah if you could even plan

for That a bit yes a building escape routes you say okay I could stop here is actually very helpful if you can do that yeah starting I don't know so should you the question is should you avoid saying that's a great question no I think it can be quite a nice things if you if you say every time you know then it gets a bit but you know often audience members he will feel quite tense ative about asking questions they feel worm-like to you are the you know The person who knows what they're talking about

so then it may be quite tentative so if you convey you know that was a really stupid question on you know what how could you possibly have thought of asking me that will be bad right so you want to convey that and maybe once or twice you will actually said did it annoy you when people do that I just sometimes feels as if it's is it so I guess some people might to kind of sort of formulaic yes yeah yes so yeah Avoiding formulaic counters some people just yes that is only better things ago I've

been doing a bit of that right so it's quite often audience members other audiences for far away have not even heard the question so repeating the question be quite a constructive thing to do it also lets you rephrase it in a way that you know is who I think you're asking this like and that gives the question which I'll say no I wasn't what I was asking at all but sometimes people Ask the most obscure things yes what do you do they don't know so if you don't have the answer they then it's better not

to flannel I mean it just it is difficult you have no clue then you just have to say they say but you know are classic is and but didn't you know Strunk and and boodle do this you know what what about their Wiswall system I never heard of them let alone the wisdom or system you just have to say that's really interesting I don't know About that can you tell me you know give me a pointer to it afterwards there's no two ways about it if you just watch all and so the people will detect

it it's better just to bite the bullet I think yeah do I practice a lot Oh before yeah so I think it's a yes it's a it's a very good idea to do a complete dry run if your talk so you won't get any questions you know in your bathroom just to check for length length is quite difficult to judge in advance People go through slides at very very different rates I have typically have quite a lot of slides but still go through them you're about to see an example of truncation it's 1029 but yes

rehearsal good good idea preferably with a friend and how it helps to do a good effect okay so there are there any key points that we at missed I'm going to just I just wanted in the last 30 seconds tell you two things one being a good audience member So you're being good audience members so this is the other side of the coin right be the person who asked the questions go and tell you how often that you you have a whole audiences that are completely lost and they're just desperately wishing that somebody would say

I haven't got a clue what you're talking about can you please send I say that again in a clearer way but none of them are brave enough to do it so you can be that person so and do that early don't Wait till very late right and then you betrayed the fact that last 20 minutes have been completely wasted if I because you so as soon as you get lost start asking questions and then some point it'll become clear that you know you're not you know that there may come a point where you have to

stop just because you aware that you're going to become that questioner who's debating the talk let's see just let me mention do not do do this Right that's really annoying and this is also really annoying this is also really annoying yes right so so they're really revealing points one by one I hate that it would annoys me right because it makes me think makes the me think the speaker didn't really trust me okay and you must stop there's what I'm about to do because when it comes to be 10:30 you're still sitting there quite politely

but when audiences expect the talk to stop they How stop listening so if you go on right then they will nothing goes in so to complete waste of time and your host gets arrested and then I've seen people you know they're five minutes over there on slide you can see in the corner there slide you know 22 out of 60 and they look at their host then they say I just got a few more slides would that be okay and the host that you know the audience are going looking at the hosting game and it

says well yes that would be Lovely so please finish and we shall so we could take a couple more questions and I'm sure we're Simon will around actually all week and I'll be back on Friday so any questions you want to wrap up or all comments and observations - from your experience then you might sound robotic I think a couple of lunches will not make you sound robotic right 51 threes might make you sad robotic so I do it a bit but but not Script every line what I want when I pause oh yes yes

I think I would so I'm done this very much today but I think if you if you think at that point in my talk I'll just pause for a moment yeah I'll look at the audience you could even say okay just make you know if it's anybody got any questions at this point and there's bond you're regrouping yes absolutely so planning some pauses I think would be good oh I contact I can't super important yes Look people in the eyes and with with a few or many Oh like I said picking a few that you

look at occasional but if you just look at one person all the time they get a bit they sort of suddenly what what is it so you just so you want to pick you know if you're gonna eyeball some particular people but but the big thing is look at your audience not at your feet not at the slide that's why you can look at your face they would feel so yes yeah Okay thank you poppy time