At the beginning of the year, I set out to communicate about a broad range of AI-related research topics and ideas. This meant more than a challenge for me, it was something that I strongly believed in and something that is desperately needed in our community. I also knew that the road ahead was going to be a bumpy one, but one that could teach me how to communicate about topics that are highly technical, extremely important, and interesting to a very “picky” community and the broader audience. The journey has just began (half way through the first year) but here are a couple things I have learned so far.
Technical stuff matters
You will be surprised by how strict and picky your audience may turn out to be, especially those that belong to scientific communities. You are not allowed to make too many mistakes here. Your articles must have depth without compromising clarity. Striking a balance between these two things are critical in a field that is as complex as is the case for Artificial Intelligence.
Avoid sensationalized headlines, keep it clean and positive
One of the reasons I started to communicate about AI-related topics is because of the abundance of sensationalized headlines out there. I am tired of the fearmongering and seeing how AI discussions are lead in the wrong directions. AI communication should not be mixed with entertainment. It should be treated with care and guided by different motives that exclude “attention seeking” and “self-gain”. The reality is that these two things spoil the conversation and are a detriment to the progress of science as a whole. If you are interested, you can find some excellent ideas about what AI-related topics matter most today in this beautifully-written article.
A good illustration, please
Again, avoid placing irrelevant images on your articles that just hurt your writing. Here is a great article discussing what type of images to avoid when writing about AI-related topics. Essentially, keep your illustrations or images consistent with your writing. This is hard because you will find that there are very little resources out there for high-quality images that are relevant to what you are communicating. Remember, the same rules that apply to your writing also apply to your illustrations: avoid exaggeration. I usually take the time to design my own art, so that way I keep it relevant to the discussion, but this takes time and creativity. You can always hire someone to design something for you. Believe me, it’s worth it!
English is secondary, precision is a priority
We can all agree that grammar is important. This is true for whatever you write. One thing I found out with my audience (mostly consisting of young academic and industry researchers) is that they are willing to give you a pass for grammar mistakes. That’s awesome because now you can just focus on what matters most: communication. Clarity is important, precision is a must.
A slightly relevant headline please — not too technical of course
On the relevancy part, avoid associating your scientific articles with cheesy headlines or what you think are buzzworthy headlines. Been there, done that! It just doesn’t work, at least not for the scientific community. They see through everything. They know when it’s BS. Your best bet is to keep it authentic and relevant but at the same avoiding jargon. There needs to be a balance between relevancy and worthiness. These requirements are fuzzy and so you need to play around with what works best for your target audience.
Education or Entertainment?
This is perhaps the toughest part of AI communication. Initially, one must get the reader excited about what they are reading. I am not talking about an abstract here. We are talking about an introduction that clearly defines the importance of the topic being discussed and how it may apply to the real world (i.e., applications). How else can you keep your audience reading through your articles? This is what I refer to as entertainment in the context of science communication. On the other side of the coin, aim for the dissemination of purely informative content — this takes care of the educational part. There are other components that matter but start with these two for now.
Keep it short
Science communication 101! Keep it short and clear. Otherwise, just share the links to the references and don’t even bother writing overviews or summarizing other people’s scientific work. There is the rest of the internet for that. (500-ish words are good enough from my experience). You should provide vivid discussions and spark curiosity in readers. Usually, I love to challenge readers by asking questions, writing calls for research, and providing extra readings. There is time for everything but don’t take too much time away from your audience.
Draft it up then cough it out
Slang-ish but very true! One thing I must make clear about AI communication or any science communication is that it is not an easy task. Like with everything else, you must have patience and perseverance. What I am trying to say is that don’t be too hard on yourself. Understand that there are plenty of people out there that are working along with you to achieve the same goals. Learn from them, draft up ideas, and ask for their feedback. This is similar to a peer-review process but with less hassle. You will be amazed by how quickly the quality of your content improves through this very simple process. One last thing, make sure to click that publish button! Take your time, no one is forcing you but understand that the world needs you now!
There are other things I learned along the way such as “know thy audience”, “diverse communication”, and “conversing with your audience.” I need to follow my own rules here and keep this article short, so I will save all of these other points for a future post. See you soon!
