[HN Gopher] How I prepare a talk for a tech conference (2022)
___________________________________________________________________
How I prepare a talk for a tech conference (2022)
Author : fanf2
Score : 91 points
Date : 2024-04-12 10:42 UTC (1 days ago)
(HTM) web link (chelseatroy.com)
(TXT) w3m dump (chelseatroy.com)
| ghaff wrote:
| I more agree than disagree with most of the points--at least for
| the typical conference presenter.
|
| The one thing I'd add (though this is more in the hands of the
| conference organizers) is that shorter (25-30 minute)
| presentations are often better than 45-50 minute ones. Some of
| this relates to what the post says at the beginning. If it's a
| bad presentation, I've wasted less time. But the other thing is
| that a presentation can cue me in to something being interesting
| more than it can teach me everything about that something.
|
| I've been seeing this as a general trend although I have
| colleagues who hate that trend and respond by trying to cram 60
| minutes of content into 25 minutes grumbling all the way.
|
| And, oh, stay on schedule. It's really rude to the next presenter
| (and your audience) if you don't. I might end up going a minute
| or 2 over if I get off-track but I feel badly when I do.
| exe34 wrote:
| I 100% blame people not staying on schedule on the organisers.
|
| I'm autistic, I'd happily cut off your microphone on time, not
| matter if you're the pope or a Nobel laureate. This is why I'm
| not allowed to chair sessions.
| kstrauser wrote:
| Oooh, can I recruit you to cut off question askers who aren't
| asking questions? Talks need someone not shy about enforcing
| that it's _question_ after the presentation, not offtopic
| rambling time.
| skeeter2020 wrote:
| Attend grad school. Questions are almost always a way to
| try and prove how smart you are.
| kstrauser wrote:
| You know it. "Well, in my experience doing this since
| 1997..." "Lemme stop you right there."
| ghaff wrote:
| Or analyst conferences, etc.
|
| Generally speaking I'm not a huge fan of public raising
| of hands type Q&A's. They turn into more of a comment
| than a question and showing off sorts of things as you
| say. Maybe a few pre-submitted questions electronically
| or just save them for later 1-on-1s.
| mooreds wrote:
| I have spoken a few times. My favorite self intro is:
|
| Slide with title "about me"
|
| Next slide: "who cares"
|
| Then I say "you came here to learn about <topic>, not me. Feel
| free to Google me if you want".
| dustincoates wrote:
| I care. I want to know that the person who is talking can back
| up the content and has experience with the subject matter.
| mooreds wrote:
| Fair. I guess I assume:
|
| * someone has read or at least scanned my speaker bio
|
| * folks want to get straight to the content
|
| * my expertise is shown by the fact I'm speaking and the
| content I'm conveying
|
| But maybe I'm wrong.
| ghaff wrote:
| > someone has read or at least scanned my speaker bio
|
| Probably not.
|
| As I wrote in another comment, a quick context-set and
| contact information in probably useful but _quick_ is the
| operative phrase especially for any background that isn 't
| directly relevant to the material at hand.
| exe34 wrote:
| > someone has read or at least scanned my speaker bio
|
| I'll do it after, if your talk catches my attention.
| vaylian wrote:
| Any claims made during the presentation, including claims of
| professional experience, can be completely made up. Besides,
| if someone was allowed to present at a conference, then that
| means that they have already been reviewed and approved by
| the conference organizers, which means that speakers do not
| need to convince you, that they know what they are talking
| about.
| jghn wrote:
| When you're going through the conference agenda, why not
| google the presenters for the talks you find interesting? Or
| read their bios?
| ghaff wrote:
| I usually flash an about me slide for about 15 seconds. In
| general, some context/web site/etc is useful. Much more than
| that is not. Even if you're a supposedly important person, I'm
| probably not that interested in your life history. And if
| you're that important, I probably already know something about
| it.
| 4hg4ufxhy wrote:
| You could just skip to the content and not do this virtue
| signaling
| exe34 wrote:
| I absolutely hate this kind of "oh I'm not going to waste
| your time, honestly, I'm really going to try to make this
| worth your while, wow look at me, I'm amazing!" talk.
|
| This is why I now love attending talks remotely. If you annoy
| me, I can mute you and focus on something else and unmute the
| stream when the next speaker comes up.
| jakderrida wrote:
| I'd add one caveat.
|
| "you came here to learn about <topic>, not me. Feel free to
| Google me if you want. Just please keep in mind that all
| teenagers make mistakes. It's not my fault you only made boring
| mistakes."
|
| To make it work, you need to deliver the last line while
| seemingly preoccupied with something presentation-related.
| Otherwise, it seems hostile. An example could be as simple as
| pretending to change settings on microphone or moving one slide
| back, then forward twice. It conveys it was intended to be
| inner monologue
| nicbou wrote:
| The author's writing style matches how she prepares her talks.
| That post is very straightforward. I agree with everything she
| wrote. I think that it also applies to other kinds of
| communication.
|
| I particularly like the bit about examples. I think that stories
| and metaphors are much more powerful than plain statements. It
| really drives the point home, especially when it's somewhat
| abstract. My all-time favourite is Steve Job's "bicycle of the
| mind".[0]
|
| [0] https://www.youtube.com/watch?v=KmuP8gsgWb8
| skeeter2020 wrote:
| They seem to have a consistent "voice", which is going to
| please some and piss off others, and that's way better than
| having nobody care.
| karaterobot wrote:
| I agree strongly with the bit about not including memes and
| jokes, to me this reads like nervous energy: here's me packing
| some entertainment into my presentation because I'm probably
| boring you. Just make it more interesting, or shorter. I also
| agree with the bit about not feeling it necessary to make eye
| contact with audience members: as an audience member, I feel like
| the speaker is about to call on me when that happens, like at one
| of those hateful interactive theater performances everyone
| dreads. Just talk to "the room" if you can.
|
| It's always amazing to me when I see a talk by someone who has
| clearly not practiced it very much. I think it's usually better
| when you've run through it so many times so that you can deliver
| it without any notes, but it's still concise and complete. I
| practice double-digit amounts of times for every talk.
|
| My tip is to record yourself practicing. Don't watch the
| recordings, no need for that. The thing is that blinking red
| circle seems to psychologically qualify as an audience (to me
| anyway) and it focuses me in on the performance part of giving a
| talk. One effect is that if I mess something up, and I'm
| recording it, I start improvising my way back on course rather
| than just starting over. There seems to be more consequences if
| you record it. I dunno, helps me.
| Aurornis wrote:
| > I agree strongly with the bit about not including memes and
| jokes, to me this reads like nervous energy: here's me packing
| some entertainment into my presentation because I'm probably
| boring you.
|
| One or two well-placed jokes can be great if they're a very
| minor part of the presentation.
|
| When someone has their slides stacked with memes or spends
| large amounts of time on vacuous entertainment content, it
| always feels disappointing.
|
| There was a period of time where our biggest local JavaScript
| conference felt like one big entertainment competition.
| Presenters were singing songs, playing guitar, showing several
| minutes of clips from TV shows, and telling jokes more than
| they were presenting anything useful. The conference was a hit
| for young people and juniors, but it became known as a big
| waste of time for everyone else.
| ricardobeat wrote:
| Guilty as charged. Are you from Brazil, by any chance?
| j7ake wrote:
| How many talks do you give a year? Seems like a waste of time
| to practice the talk that many times.
|
| what works for me and is more time efficient is to get the
| introduction polished, then have a clear take away at the end.
| The middle stuff is better to go impromptu.
| karaterobot wrote:
| When I have a talk to give I can't think of any better use of
| my time than getting it right. If you don't need to practice,
| good on you--a lot of people who don't probably should,
| though.
| Loughla wrote:
| The number of people who believe they are good at impromptu
| speech is WAAAAYYYY larger than those who are actually good
| impromptu speech.
|
| In general, the best speakers are those who a) know the
| material inside and out, and b) have practiced shit loads.
| j7ake wrote:
| Maybe because I am in research, but the middle parts of
| research talks are talking about their own research. If you
| don't know that material enough to talk about it in detail
| impromptu then you have no business giving talks about that
| topic.
|
| Intros and ends though I agree require careful thought to
| appeal to the right audience and give the right message.
| jghn wrote:
| I agree with you but that doesn't necessarily invalidate
| GP's point. It all depends on how familiar one is with the
| content.
|
| For one off talks, or decks that one doesn't use often, yes
| you're 100% right. When I'm in that situation I rehearse
| the talk over and over and over again to the point where I
| no longer need notes. My goal is to get to a point where I
| neither sound like I'm reading from a script, nor would I
| get thrown off if I get interrupted.
|
| However for material that I've presented a bunch of times,
| I no longer need to rehearse. There was a period of my life
| where I could on the fly give a talk on a particular topic
| at a moment's notice. Ideally I'd have a bit of time to
| customize some bits to the specific audience/venue, and
| sure I'd rehearse those. But the bulk of it I could stitch
| together content in just about any order on the fly.
| dmckeon wrote:
| Pet peeve: During Q&A, the presenter should repeat the question.
| Even if the asker has a microphone, they are often not clearly
| audible. Repeating the question, or a brief version of it, lets
| the audiences, both present and via video, hear it clearly, and
| shows that the presenter heard and understood it before they
| answer it.
| samatman wrote:
| Hard agree. The OP is full of good advice, but this is the
| single most actionable thing I could say about tech talks.
| Maybe 10% of presenters do this; everyone should.
|
| Another advantage is that the speaker doesn't always get the
| question. Either can't hear it well, or the question itself is
| weird, complex, off-topic, or some variation on not-even-wrong.
| So they end up answering something else. If they
| repeat/summarize the question first, at least they're answering
| _a_ question, and when this happens, the question which gets
| answered is often better than the one which was asked.
| semitones wrote:
| > I also think legitimacy by proxy is gross
|
| I disagree, legitimacy by proxy is a very useful _initial_
| signal. Someone who works on the gRPC team at a large API
| distributor is more _likely_ to have better advice on best
| practices for protobufs, than someone who works on optimizing
| ray-tracers for TempleOS. Relying on proxies, and likelihoods, is
| crucial for the timely processing of information. We heavily rely
| on at least some level of trust - otherwise we would have to
| completely verify every single thing that anyone says, before we
| can begin to do anything with the information.
| jasonlotito wrote:
| As someone who has given numerous talks at conferences, 100% yes.
| This is all good advice.
|
| "What new skills do I want my audience to have, and know how to
| use, when they leave this room?"
|
| That's effectively the same thing as the way I look at it. "What
| actionable thing do I want people to take away from this talk?"
|
| I will add: Practice! You should have given your talk multiple
| times before you ever give it publicly, and you should have
| watched it. That means recording yourself and watching yourself
| and listening to yourself. Yes, this takes time. But it's
| important.
|
| Also: do NOT plan to do live demos or live coding. You can easily
| record these and present them and know they work. They are
| edited, there are no mistakes, and things happen. You can even
| make the joke that it was "live coding when you recorded it."
|
| Yes, if you have to answer a question with some live coding,
| fine. But it doesn't need to be live. No one wants to watch you
| typo stuff, complain about the network, etc. Just record it. If
| "recording it takes too much time" then doing it live is going to
| suck.
| ghaff wrote:
| There are demo gods out there. Kelsey Hightower comes to mind.
| But, yeah, generally speaking, it's a high-wire stunt you
| probably shouldn't attempt to pull off--especially without a
| reliable smooth plan B.
|
| As someone who has presented a LOT, practice is useful. I
| always do at least a bunch of mental run-throughs. That said, a
| _real_ run-through, which I don 't always do, results in a
| better Take 2. So, yeah, do a real rehearsal when you can.
|
| (I won't practice dozens of time though. At peak I presented a
| lot and practicing for a given presentation that much would
| have been impracticable.)
| alexashka wrote:
| All else will be forgiven if you have something of genuine value
| to present.
|
| Remember - Stephen Hawking presented his findings. It wasn't his
| awesome presentation skills.
| kstrauser wrote:
| I was walking over to the conference room where I'd agreed to
| demo a project at a huge convention. I saw other people with
| their laptops and card tables. Visitors would drop by and ask
| questions. Demonstrators would answer them and show how they
| worked. Piece of cake.
|
| One of my friends casually mentioned that the A/V team had our
| podium set up. Podium? For what? So that you'll have some place
| to set your laptop when you're presenting the project to the
| large room we booked!
|
| That's how and when I found out I was giving a tech talk. It went
| well. Honestly, that's about as much advance notice as I like.
| You can't worry or procrastinate too much when you've got about
| 20 minutes to get ready.
| skeeter2020 wrote:
| The best piece of advice here is meta: whatever you do make bold
| choices. I don't like a lot of the speaker's decisions, and
| (aside from the memes and gifs where I do strongly agree) go the
| other way, and that's OK, they're still going to engage me. The
| number one job of the speaker is to stake out a position that
| might be contriversial or not universally accepted, and drive
| deeper thought, discussion, conversation and progress.
|
| And practice. Please practice. That's what makes the timing
| impactful & the jokes work.
| mvkel wrote:
| The advice feels solid, well at least it speaks to Chelsea's
| personal approach to making a presentation. Authenticity counts
| for a lot.
|
| But those slides; the walls of text. If I'm reading what's on the
| slide, I'm not listening to what you're saying. Worse, I'm trying
| to do both, so neither is leaving an imprint in my brain.
|
| Text-heavy slides are great if I'm emailing a deck to someone,
| but for presenting, 5-6 words max per slide.
| ghaff wrote:
| The answer is ideally that you have a presentation deck and a
| linked article/blog but that's asking a lot of presenters who
| don't have the article/blog published somewhere anyway.
| cmgriffing wrote:
| I think some of the bits of advice here apply more to "soft"
| topics like tech debt, etc. Things that feel more like
| storytelling rather than informational.
|
| Technical presentations have a different set of "rules".
| com wrote:
| Which rules do you think are missing and which shouldn't be do
| e for tech presentations?
| pyuser583 wrote:
| Lots of different views on how to do a talk. This goes against
| some mainstream advice such as: establish your credentials.
|
| But rule number 1 for any talk is "know your audience." Since the
| audience is tech people who care about tech issues, talking about
| tech is a better way to establish your credentials than listing
| your CV.
|
| But then again, how the people in the audience know you aren't
| just talking smartly?
|
| My favorite tech talks are by people who jump straight into the
| tech issues without introduction. But I'm fully aware there's a
| chance they could be straight up bullshitting.
| jghn wrote:
| > some mainstream advice such as: establish your credentials.
|
| I've found that this is domain specific. You see it more for
| tech conferences for instance, but less so in biotech
| conferences
___________________________________________________________________
(page generated 2024-04-13 23:00 UTC)