AN UPDATE:
I’ve decided to start including the links to my favourite things from the week at the top of the emails.
The more I’ve got into the writing of these weekly emails (or ‘idea cultivations’, as I like to pretentiously call them), the longer my writing has become. That’s been fun and all, but one of my objectives when I started writing these was to share the best things that I had come across on the internet. This is much less effective at the bottom of an email, after a 1,000-word article.
And while I try my best to write good articles each week, there are people on the internet far smarter, more articulate and probably with more free time than me, so it makes sense to promote the best work I can find.
I will still share my writing from each week, and my plans for the next few months are series on:
Should we become “content creators”? If so, how?
Becoming a data scientist (from any background)
as well as mix of reflections, thoughts about machine learning and probability some productivity tips.
THIS WEEK’S LINKS
Machine learning in healthcare
This week I’ve been thinking about how I can use my machine learning skills to support the fight against COVID. Here are two great articles, which outline potential approaches and point out that most models perform pretty poorly:
Personal
After many months of hard work, a paper I’ve been working on was published in the British Medical Journal. The paper looks at the many research papers claiming their AI algorithm has comparable or better performance than doctors, and concludes that the claims often over-reach and the research methodology is generally poor:
I was interviewed on the YouTube channel of Ollie Burton about machine learning in healthcare, approaches to learning medicine and alternative careers:
TO CREATE OR NOT TO CREATE? (ARTICLE)
The idea of ‘creating content’ has been in the periphery of my mind for quite a while. Broadly, my feelings have been mixed; I see the advantages, but I also see some drawbacks. I see value in “being a creator, not a consumer”, but I’m not convinced it’s not for everyone.
Today I’m going to try and hash out some of the arguments that have been bouncing around in my mind for a while.
As I quick definition, I would consider a “content creator” as someone who produces material (blogs, videos, tweets, artwork, etc) that is shared publicly on the internet, often with the intention of building a following.
For me, there are three main reasons to create content; gaining transferrable skills, the benefit of public accountability and scalability of impact. On the reverse, there are three main reasons not to create content; losing time for deep work, losing focus and authenticity in popularity and potential unforeseen negative repercussions. This article will take a deep-ish dive into each of these.
Reasons to create
Gaining transferrable skills
The cornerstones of ‘content creation’ typically involve writing, taking photos and making videos. All of these involve a fair degree of skill and creating content can be a great way to develop them. There are many ways in which such skills can be advantageous in other areas. I believe writing well is largely synonymous with thinking well, which is an ultimate meta-skill. Being able to present things aesthetically well, through photos or videos, will be useful anytime you want to create a presentation, or promote something.
Looking at it through this lens was a gamechanger when I was thinking about starting a YouTube channel. I was initially hesitant to make videos, as I wasn’t sure I had anything valuable to say and felt my filming and editing would be too poor. However, when I switched to viewing it through the lens of learning a new skill, it felt much easier to accept poor quality to begin with and to view it as a fun side project.
The benefits of public accountability
There is a lot to be said for unbiased feedback on your work, and creating and sharing content publicly can be a great source of it. By sharing work in public, you gain a metric of how well it resonates with others, which is a fair proxy of how valuable your work is.
This kind of unbiased feedback can be hard to come across in typical day-to-day life. If you show your work to those around you, they will typically have some sort of relationship with you that may have the objectivity of their feedback. There is nothing more objective than whether a stranger from the internet decides to spend time consuming your content, or not to.
A couple of caveats: Firstly, your work must be discovered by sufficient numbers to actually get this feedback. Secondly, there are times when attention may not be the best metric of value. Some would argue the Kardashians are an example of this. I think it depends on the type of value you are considering.
Scalability of impact
Traditionally there has been a bottleneck on the number of people your work can reach; only so many people may have access to your book or be able to hear you talk at a venue. The internet has of course changed that; anything you make and share could in theory be read by anyone, including people who don’t yet exist.
This increase in scale increases the impact that content creation can have. I’ve felt acutely aware of this when teaching medical students. Sometimes, I’ll spend a couple of hours preparing, and a couple more hours delivering, some teaching for a group of, let’s say, five students. Even if the teaching is amazing, and sticks in their minds for the rest of their lives, the benefit is limited to those five students. In comparison, if I spend the same four hours making a great educational video and share it on the internet, there’s a possibility it would be seen by thousands.
For me, this is a strong argument for looking to convert what you do in the real world into some sort of digital content.
Other reasons to make content
Enjoyment is another big factor for me; there’s something exciting about trying to create something of value, and about putting yourself out there on the internet. Other factors could include a creative outlet, a pursuit of influence and fame or a source of alternative outcome, although these are less the case for me.
Reasons not to create
Documenting, but not creating (and losing time for deep work)
There’s a mantra in the content-creation space which is “document, don’t create”. The idea is that you shouldn’t aim to create new pieces of content, but rather your content should be a documentation of things you are doing anyway.
And while I believe this is a really helpful way to view things, there’s no getting away from the fact that it takes time to document things. One fear in the back of my mind is that by spending more time documenting, I will have less time to work on other projects and may ultimately reduce the amount I can achieve.
Elon Musk can be an example of this, at an extreme. He has worked intensely towards bringing technologies that could change the ways we live. But he has spent almost no time “documenting” that process, and I believe it would take away from his work. And because of his success, there are a lot of other people now documenting his work.
The main counterargument to this is the potential scalability of sharing your documentation, that may eclipse the other work that you do. Eliezer Yudkowsky shared a really interesting piece about this (although I couldn’t find it! – I will add it here if I can later).
Losing focus and authenticity in popularity
The flip side of using popularity and attention as a metric of value is that it may steer you in a different direction to the one ‘most authentic’ to you. The ideal scenario is when what you want to make and what is popular perfectly aligns, but it may not always be the case. The best defences against ‘losing authenticity’ are probably clear insight into what it is that you want, and maintaining a focus on content rather than audience numbers.
I want to demonstrate what I mean by this with a slightly contrived example.
Let’s say I’m a researching applications of machine learning in neuroscience and want to create resources for the research community. Given the niche-ness of this field, there’s only so much ‘traction’ such work could gain (although this doesn’t limit the value that it could bring).
Then, one day, I make a video about AI robots in movies, which becomes extremely popular. This sudden spike in interest impels me to make another video, about AI predictions from books, which again is popular.
And while this new style of videos is very popular and may open doors to me as someone explaining science to a broader audience, it may also take some of my focus away from my original research focus. Maybe I start getting invited on the BBC or radio shows and look to keep making videos of the same style. The tricky thing here is that such opportunities are really hard to turn down once they arise.
And let’s say that, in a hypothetical alternative universe where I monomaniacally devote all my focus on my research interest, I happen to make a landmark discovery that leads to development of a brain-machine interface. This may never have happened had I taken the alternative route.
Now of course this is a somewhat artificial example, and you can make a strong argument that the ‘alternative path’ (of making popular content) just adds a different kind of value, this is a fear that remains in the back of my mind as I consider whether or not to pursue ‘content creation’.
Unforeseen negative repercussions
A flip side of the scalability argument is the scaled-up potential for an unforeseen negative result from putting more of yourself out there on the internet.
An abstract fear in my mind is that, let’s say, I decide to act a bit goofy in a video I make. Then, by chance, a potential employer that I’m interviewing with stumbles across that video, sees my acting goofy, and assumes that I’m not serious enough for the role. It’s hard to regulate who sees what you put out on the internet, which is both fantastic but also a little scary.
I worry that, even with the best of intentions, there is always the potential for things to backfire in a way you can’t necessarily predict. For example, I’m aware of at least a few cases where a rogue tweet or a side comment made on a livestream have ended up costing people jobs.
My main thoughts to counter this are:
People give you much less attention than you think, and the chances of e.g. an employer watching your YouTube videos or reading your blog are extremely low
You can deliberately keep certain personal stuff out of the content you make
Going forwards
In the course of writing this, I’ve come to strengthen my view that creating content is something worth pursuing. I feel the pros had more weight and were more concrete that the cons. I’m still not fully-invested, but I’m more ready to devote a moderate portion of my time.
In that vein, I will be committing to, for the whole of 2020, to:
Writing an article every single week
Creating a YouTube video every single week
I believe this is not so big a time commitment that it will massively detract from other tasks, and I think it can be a fun side project.
I’m also deliberately not creating many constraints around these two commitments, such as content or format. I want to give myself the freedom to experiment and deviate.
By the end of 2020, I should have a clearer idea of whether this is for me or not, and will either double-down or move onto something else.
In the best case, I’ll have found a new I’m passionate about and perhaps built an audience. In the worst case, I’ll know that It’s not for me, will have a portfolio of content to look back on and will have concretely developed better writing and video editing skills.
Hashing out the negatives has been helpful as I will bear them in mind as I progress on this journey, and keep working to minimise them.
Let’s see what happens! Feel free to subscribe here to follow the journey.