I’ve previously written to call for more writers in the data space, but, since I wrote that piece, there now seems to be an abundance - too much to realistically consume every week.
What inspired me to write this post was a call on LinkedIn, to see if anyone wanted to join an accountability group about writing and consistently do so over a long period. There are many substacks that I’m subscribed to, which have one or a handful of posts over many months or years. This is not a problem in and of itself, but it is as if the writer hasn’t succeeded in achieving what they wanted to.
If this call is going out, is there a problem with people writing regularly? I wonder why. The truth is, I can only guess why other people struggle…. time, priority, distraction, lack of inspiration are all things I’ve struggled with.
is not going to write a post this week. After 105 weeks continuously posting - the end of a great run 🫡. Incidentally, the first post of this run, “The more the merrier”, was the gauntlet I picked up to start this substack. What is it that enables someone to write consistently for a long time (assuming they want this)?The LinkedIn post above proposed elimination if someone in the group didn’t write for two weeks. While motivation is needed, I’m not sure if punishment is the answer - it gets a bit like the Hunger Games. I feel like it may be better if everyone in the group had some kind of counter of days since last post, and, once it gets to 9, others could reach out to see if everything is OK or if they need some help. I’m down to help either way, though, for anyone else who struggles with this.
So, to recap on the problem space: we have plenty of writers now, but many that struggle to keep it up or don’t reach their goals in writing.
I can only comment with perspective from my own process - what may be true for me may not be for others. By the same logic, what is true for me will be true for some others, too.
I feel like the three dimensions we should consider when writing are what, how and when we should write - especially in the era of AI. My feelings on this have got a bit stronger since I started writing. At that time, the idea of AI-generated content wasn’t even something people really thought about. It was perfectly valid to write standard copy about useful topics and have this kind of content become well-read. What I would say today is: if anyone could have written the content then, AI definitely could have, too. If the content is standard, generic or impersonal, then AI reduces the value of it, as you can now get that kind of content on demand.
In response to the devaluation of this type of content by AI, I feel we should write as humans. What do I mean by this… of course we write as humans, do we even have a choice? What I mean by this is that your writing should be as unique as you are. Don’t suppress any part of you (not more than you would with a friend): emotions, preference, humour, hypothetical thoughts… it’s what makes you you, and is what is not AI (at least for now). That doesn’t mean you can’t cover technical or factual topics, it should just be in your own way. When someone reads what you write, especially for someone that knows you, it should be as if you were speaking to them. I’ve met some of the folks whose substacks I read, and when I read their posts, I can hear it in their voices.
The advantage of embracing your own voice is not only that it makes your writing unique, but it’s also more efficient. If you’re not trying to polish your thoughts into someone else’s, if it’s mostly unfiltered and raw, with editing only for coherence and grammar, it takes a lot less time.
I think you shouldn’t be afraid to do something different, too. Last week, I took a detour from prose just because I felt like it, just because I was in the seasonal mood. I hadn’t done so since I was about 14 or 15, and it probably showed! Don’t be afraid to diversify in more than form, too - I write about anything I’m interested in or care about. Yes I’m a data person and I think about data stuff a lot, but I occasionally think about other things, too. Writing about all the rest makes all of your combined writing uniquely your voice, too.
I mentioned when we write - by this I include the circumstances, too. I’m not someone who can set a two hour slot every week and generate a post. There are times in the week when I can write very easily, and times when it’s more or less impossible. I really only write anything in the morning when I’ve started writing a post the night before. My best hours to write are probably between 10pm and 2am. I don’t know if that’s because the house is quiet, my mind is quiet and undistracted, some kind of circadian rhythm thing or a combination of these and more. It’s just who I am, and I’ve learned not to fight it. This is where I feel writing is really a physiological thing as much as it is a mental and psychological one.
The deadline of Wednesday every week for 100 weeks helped me find a rhythm, and it really is a physiological rhythm now. My brain expects to produce a post every week - it’s a level of fitness that you can reach, like practicing shooting from outside the “D”.
I’m really not a natural writer - my natural inclination is towards STEM. I always find it incredible how quickly
writes great content when I ask, but he’s been doing it throughout his whole education. Our roles as founders are the reverse of what our backgrounds would suggest. I would avoid written work by any means necessary in my education, always preferring more weight on exams. I’m saying this because I think this level of fitness is something many can achieve. I probably have a very low natural propensity for writing - if I can do it, you can do it.It does take me longer than others who are naturally inclined towards it. I often start a post a few days or longer before posting it. Sometimes weeks, sometimes months. I have many posts in progress - some just ideas jotted down as post titles, some that are nearly complete. Forcing yourself to finish a post isn’t always the best way. I’ve found weeks where I’ve got pretty far with a post, then hit a wall or got distracted by a new topic and ended up publishing something completely different. I would advise not fighting this where possible. You probably don’t want to have tens of highly progressed but unfinished posts, but having a few is fine and perhaps even healthy. It allows you the flexibility to respond to things as they happen and respond to your own levels of focus and attention.
Happy New Year for Monday 🥳! Thanks for sticking with me for two years.
I’m optimistic about the new year ahead. For a long while now there has been a story of doom and gloom in the startup world.
’s post here explains why and its relation to ZIRP. There is hope that high interest rates will lower next year and, in turn, have a positive effect on VC funding.2024 was gleefully predicted, by some, to be the year many startups go under, as they can’t find new funding - there are almost certainly some that will and others that got funded during the frothy heights of 2020 and 2021 that shouldn’t have been. There are many others that have promise and need more time=money to prove their value or find a reasonable exit. I think with interest rates increasingly predicted to decline over the next year, it’s not reasonable to assume that they will stay high for a long time. It’s also reasonable to assume that they probably won’t be zero again for a long time. It feels like they will likely be around the 3 to 4% range for a long time, as central banks try to balance inflation and growth - a medium interest rate era: MIRE.
There will likely be some threshold interest rate, which, if crossed, triggers a great deal of funding to go back towards VC to find returns not possible elsewhere. I’m not sure about what exactly that is, but as interest rates decline, startup valuations increase > as valuations increase, portfolio returns increase > if overall VC portfolio returns become higher than other investment returns linked to interest rates, then funds will begin to flow back towards VC.
I also think that this is the year that we see AI become less hyped and more normal. People will begin to see the benefits in many places: MS Office and other business applications, their smartphone, their car…
The AI arms race has been pretty furious this year, too - I can’t see how we get to the end of 2024 without GPT-4 being dethroned by a model from OpenAI or Google. The progress of open-source and methods of using AI have come a long way, too. I think there will be open-source recipes to get to near (enough) GPT-4 performance at lower cost and latency, available by the end of next year. The likes of Mistral and Anthropic could well deliver a GPT-4 level competitor also, at a fraction of the cost and with improved latency. Cost per token has only one way to go in 2024 - down.
I also think there will be more consolidation in the MDS space - yes, some will be forced and acqui-hires, but some will genuinely be “better together” stories - bundling to win.