← Achraf ASH

12 Things I want to try in 2022

Achraf Ait Sidi Hammou

The problem I have with goals is that they require a clear vision of where you want to go.

I don’t have that.

Plus, goals usually reinforces something that you already have: an idea, a plan, a project, a habit.

Right now, I want the opposite.

I want to try as many things as possible, but of many different kinds (apps, content, community, service).

So here they are, the 12 things I want to try for the remaining 4 months of 2022 (and probably the year after because it’s way too much):

1. Become an ML research influencer on Linkedin

I kinda started this earlier this year and wanted to make it the highlight of this year.

But I quickly realized how much of a treadmill it was and impostor syndrom hit me hard.

However, the few ideas I shared seemed to have resonated. A few people reached out to ask me questions about ML. I kind of became the reference in my close N+1 network.

Now I want to take it more lightly - like Twitter - and just share rough ideas and projects I’m working on.

I also want to build some kind of brand. I was thinking of something like being the guy who always talk about ML and yet always argues against the use of it (rule #1: don’t use ML).

2. Create serendipity chats with industry experts wanting to know more about AI (from my Linkedin presence)

As I said above, I want to get out of my bubble. And the one thing that both freaks me out and excites me the most is talking with industry experts about ways to use ML to solve their problems (or even disrupt the market, like Doctrine.fr did).

I’m positive that if I share the right content - not too technical, but enough to seem legit - and optimize my profile, I can get people to reach out to have a chat around their problems (or I could just cold DM them directly).

3. Selling info products (templates, courses, code/notebooks)

Nothing crazy, but I think the process of selling a product for the first time kind of breaks a ceiling.

I don’t know about you, but the idea of asking my audience to buy something I’ve done is terrifying (not rational at all!).

I want to do it just once to go through the process of:

4. Producthunt launches

I kind of did a Producthunt launch but it was just to test the platform.

I want to do a state-of-the-art launch:

I definitely think that launches are overrated but still an essential skill to develop.

5. Selling a product on Microacquire.com

You know the meme about startups?

Nobody knows what step 4 is so let’s skip it.

More seriously, I think that selling a micro SaaS is great for financial reasons (duh) but also to learn the importance of setting up your business correctly to get acquired (aka making yourself redundant). That means building a marketing machine that brings customer without your input and a product stable and maintainable enough to be taken over.

I’ll probably sell my first project that reaches 50-100€ MRR.

6. Creating a wellness community/forum (and monetize it?)

I’ve always been a wellness and health optimization nerd. The type of person who can binge watch every Huberman Lab podcasts for hours.

I don’t really have any idea of a product to build for that audience, but something I haven’t found is simply a good place to hangout for more science-backed wellness practices.

There are not a ton of communities around wellness/health optimization and most of them are full of gym bros talking about nootropics all day 😑

A website that I’ve been loving lately is theproofwellness.com. They have some amazing interviews. That’s the type of thing I’m looking for.

7. Productized service / system-based business

Productized-service agencies are all the rage in 2022. Especially with DesignJoy.co making waves on Twitter - at first for good reason (awesome numbers) but then for bad ones (under delivering or even ghosting clients).

I love systems and framework. I’m always in awe when I listen to interviews of entrepreneurs like Andrew Wilkinson who are able to completely systematize a business and get themselves out of the day-to-day operation.

I love that it’s a pure meta game. It’s all about getting the numbers right (margins, acquisition, etc), finding the right offer, doing good marketing, and hiring the right people.

I think there is a LOT of room to build a DesignJoy for DevRel.

Here is my plan for it:

  1. cold email fresh startups building a product for developers with a great message full of value (teach them about SEO, how they can do it, caveats about content for devs, etc) and offer them to take care of content marketing while they get their product right;
  2. manage a few clients (2-3) with an offer like $2k/month for 1 blog post per week and a weekly (actionable) report (numbers, results, suggestions);
  3. find good enough people writing for fun on dev.to, medium.com and hackernoon.com, and hire them (as contractors) to write the blog posts - make some edits to make sure the quality is constant;
  4. build some systems to automate much of the work: marketing engine with SEO and socials, bots to find young promising writers, wikis on how to write effective content;
  5. make yourself redundant;
  6. sell the business.

Honestly, I’m really tempted to do it, but I’m really afraid that I might have to work 100 hour weeks and get stuck because I have no time left to go hire people.

Maybe I can hire people from day 1?

To be continued…

8. Create Daily Muslim a newsletter for Muslims in tech

The best hack when you are addicted to the game (aka workaholic) is to find a business that lets you take care of another important area of your life.

For me that’s religion.

James Clear said:

A project without a goal is a hobby. A goal without a project is a dream.

I’ve always had religious goals but never made a single step towards them.

But a few days ago, I was reading Ryan Holiday’s Daily Stoic and realized that we Muslims already have that type of lessons, from the Prophet (SWS).

So I want to create the same type of newsletter except it’s lessons from hadiths.

The format could be something like: 1 story, 1 actionable lesson adapted for the modern world, 1 challenge to take action this week.

9. Create a Deep Learning framework and ecosystem for Pytorch like Vercel and NextJS for React

Seeing how easy it is to build a web app in 2022, I can’t but think that the future is similar for deep learning.

Picture this:

  1. bootstrap a project from a microservice template,
  2. use a framework to quickly define your data pipeline and model using Pytorch,
  3. push your changes to Github,
  4. configure your project on the infra-as-a-service (Vercel equivalent),
  5. your API gets built and redeployed everytime you push a new commit.

10. Build interesting Deep Learning research projects, and get paid to write blog posts for tools I use (wandb, banana.dev, hugging face)

The past few months, I’ve done DevRel for the first time and it was really fun!

I’m interested by many research topics but rarely have an outcome in mind (using it in a product).

So maybe the outcome could be a blog post to share the results.

I want to rethink the way ML research is done, by making it more accessible.

That means:

I don’t want to think it soo much so a good format (to force me to follow up) would be to pitch a research idea to a company like weights and biases, banana.dev or hugging face to 1) give me some free credits and 2) publish a detailed article on their blog against a small compensation (a few hundred dollars).

11. Build an execution-driven devtool bootstrapped startup: win at SEO, great design, changelog, opinionated, great market but saturated. Do some research to find the most searched alternative and great market.

For years I’ve dreamed of finding a project that is 100% about execution.

Things I want:

Examples: savvycal.com, linear.app, loops.so

Ideas: Metabase.com alternative, Sendgrid.com alternative, RSS Reader redesigned from the ground up.

12. Growing on YouTube and/or Twitch

Earlier this year, I tried streaming on Twitch and YouTube. I didn’t put that much effort into it but I got some good reception.

The first time I got to chat live with viewers on Twitch was definitely awesome and weird in a good way.

But that shit is hard. You have to sit down for 3-4 hours straight and be presentable. Most of the time I had to call it a day after finishing.


changelog