One of mine, which I put into practice after attending a JavaScript-focused conference in 2018: Don't attend conferences that focus excessively on latest-thing social issues or identity politics.
Conference rules/terms/bylaws that are longer than a Life Insurance contract are also a personal red flag.
Yes, I work in machine learning and some (many) of the conferences are very scary politics-wise.
Personally I think it's better to focus on conferences that are smaller and more niche and don't have the baggage that comes with getting big and famous. Any time spent on politics is time not spend on actual research discussion
Very nice. I disagree with not going to recorded ones. When is the time to watch recorded talks: never. So it is now.
And the 5 minutes talk 17 min discussion could work, but it really depends on the company. If head of e. g. darpa gives instead of a 1h keynote a 15 min teaser and then all 1000 participants can start discussing, it might not really work (as other situations as well).
Beyond the scenario you describe, I have some apprehension around audience questions in these venues, as the questions are often enough digressive or part of some dominance/status game.
For a speaker to aim for presenting for 25% of the time and fielding questions for the remaining 75% strikes me as lazy or a dodge in case you have no idea how to effectively present your information.
Totally agree with you on the recorded talks, BTW!
>If you’re presenting, leave more time for questions than you’re supposed to.
My personal corollary to that: when you're preparing a talk, focus less on trying to get all your findings across and more on what questions you want the audience to ask you. The least interesting talks are those that are unfocused and overwhelm the audience with a firehose of information. The most engaging talks are those that guide the audience to ask interesting questions.
I agree with all the rest of the post but interestingly I would say the exact opposite for each of the first three points.
> Don’t attend any talk that’s being recorded
I know myself. I've tried watching those back but it never happens. At best one or two talks if I'm particularly curious about one thing and there's no info about it elsewhere.
Which makes me think talks may just not be the best form. I guess I like them for the combination of entertainment while getting new info.
> You’ll get all the same info in 10% the time by skipping through the video later
That's the problem with video, though: it's almost impossible to skim. I'd be curious if someone quizzed you on those talks after watching them in real-time versus with skipping. Sure, 10% and "all the same info" are both hyperboles and are not to be taken literally, but managing 80% of the info in double that time (20%) would surprise me.
> Caveat: if attending the talk will let you meet the speaker (usually it won’t)
What conferences do you go to where the speaker is locked away after the talk? This confuses me. Or do you mean the line will be too long anyway and there is no point trying?
>managing 80% of the info in double that time (20%) would surprise me.
This isn't particularly difficult. If you skip the pleasantries at the start and end of a talk and use a playback rate of 4x, you'll more or less get that 20% duration. 4x is quite understandable (once you're used to it) given how slowly people giving talks tend to talk. Though if a talk is really dense in terms of content, you might have to go slower just to give your brain a chance to digest the information. Most tech talks aren't that dense, however.
Honestly this depends on the conference, but if it's the kind of thing where it's a CV filler mill you won't learn anything from either the talk or reading the paper. And then there are talks which are impressive because the authors did a lot of work - again you don't learn anything because the solution is obvious, just takes a lot of effort. In those cases I agree it's better to shmooze.
As someone who has given a talk, here and there, one of my favorite things to do (and attend), is small, lightly-attended panel discussions, with a loose format, and the projector available to everyone at the table, via an HDMI switch.
They are best when the panel encourages the audience to participate, and have interesting people sitting at the table. These may not be jargonauts, but people that don't intimidate folks for asking questions, and speak in the vernacular. It helps if the folks at the table have good interpersonal chemistry.
One of my favorite sessions (that I attended -not spoke at), was at a MacHack conference (yeah...old), given by a couple of guys from the Apple "Blue Meanies" team. They were almost like a comedy duo, and gave great information, in a relaxing manner.
MacHack was great. It was a tremendous opportunity to interact directly with Apple geeks.
The standard Apple "One Guy/Gal With Xcode on Presentation Mode" talks are great fodder for after-conference review, but B O R I N G to attend. The production quality is usually great, but I never seem to get much out of the one I attend. I have to review it, in order to understand it.
I tend not to attend any talk where I recognize the name of the speaker's company. Two-pizza teams from giant companies often have talks that either were only prototypes, or were actually tire fires in production, or will soon be shuttered. They can't give you the straight talk because it might embarrass the company/team.
If it's a big conference and the speaker is from a no-name or I don't recognize them, I expect the talk was accepted because it's actually an interesting talk and not just coming from a big name company or a sponsor. I also get a lot more out of non-technical talks.
At the academic ones, it's mostly cliques of grad students from the same lab, and they maybe socialize with grad students from labs where their advisors used to be. I've never seen much serendipity happening.
The tendency is even stronger at tech meetups and conferences. People use them to meet up with friends they rarely see, or they just come with friends from work.
As few connections as I've made on technical Discords, I've made fewer at conferences. Go if you have people to go with, just watch the videos otherwise.
So I'd love some advice or help I've never really seen the appeal of attending conferences.
It seems like there is so much information presented so rapidly that I end up forgetting most of it pretty quickly and am only able to focus on a few things. I don't drink or really party so that doesn't really do anything for me.
It seems that if I want to learn about something the best way for me is to sit down and read about it for a while. Not to mention half of it seems to be marketing drivel depending on the conference.
So HN what am I missing? How do I make a conference useful?
The key is meeting people who are interested in similar things, or working on similar problems.
That's it - that's the entirety of it.
The dirty secret is conferences would be much better if they simply had no sessions scheduled at all, and locked everyone in the ballroom for the entire day.
I've run a very small (~20) conference over the last 15 yrs and this is basically the plan. We even had it on a cruise ship one year. The friendships, and collaborations that have ensued are a testament to its effectiveness.
(To be honest we do have a few short talks and an agenda to guide the discussions, but these are mostly requirements of the funding that we've found)
> The dirty secret is conferences would be much better if they simply had no sessions scheduled at all, and locked everyone in the ballroom for the entire day.
I've been to conferences where I simply forgot to attend any talks at all, simply because I was too busy/having fun chatting to people.
A lot of the more "professional" conferences have stands just for paying sponsors though, and usually not a lot of them. They tend to be a lot less lively than open source conferences and the like which allow any ol' open source project to set up a stand. I like stands because they're a good conversation starter, and in many ways also like a presentation yet a lot more informal.
Unfortunately all of that died with COVID in my area, and hasn't really started back up since :-(
Corollary to this is that Conferences are judged not by their overt content, but by their curation of the un-conferences / birds-of-a-feather meetings that they facilitate.
You can walk up to and pester really smart people and have a conversation with them. If you have ideas you want to discuss with your peers, this is a great place. Or if you want to hire somebody, or keep your ears out for a better job, it's a great place to network. Give yourself an objective or two over the period of the con, and consider the rest of the time just a way to collect ideas, and maybe sightsee if you get bored. You may also prefer unconferences.
Never go to the sessions at a commercial conference. Academic conferences may be different. The vendor fair and social interactions are the only useful part of a conference.
Meeting people is a big part of it. Even if you're not very social, I found attending conferences to be great for understanding what people are doing now, getting a feel for the personalities in my field, and making connections and just learning.
When I used to go to more academic conferences, I mostly went to smaller tight-knit ones where most people knew each other. I found that much more interesting than the big ones with thousands of people which definitely feel more anonymous and take way more work to meet people in, and end up being closer to what you'd get from attending something online.
Tldr, try going to a ~100 person niche conference in your specific field, or at least one that has tracks that emulate that
> Tldr, try going to a ~100 person niche conference in your specific field,
Where does one find conferences that small? Everything I've seen (is|seems to be) for large-ish conferences? Large-ish being defined as the ones with the "anonymous feel".
Conference rules/terms/bylaws that are longer than a Life Insurance contract are also a personal red flag.