# What to Focus on, remainder of 2021 ### Questions to answer... * What is the problem that you are trying to solve? * What is the most important thing? * What could be a third option? * ### Background **Why are you feeling overwhelmed?** It is hard to know what to pursue. I am trying to catch people who are *so* smart and have 5-10 years more experience. It also isn't clear that my actions and goals are in line with what my new team (AI eng) needs **What changed? Why was data science different?** * I don't believe that Data Science had quite as much technical talent, and I was able to find a *niche* that I could learn about that others didn't have a background in. DS was so broad (touching on math, computer science, business, psychology, experimental analysis, etc...). * I had *systems* and *routines* that worked. What worked well *there* may not continue to work well *here*. Now is a perfect time to update these routines. In fairness to you, it is hard to know what a new routine should look like until you have been in the thick of things for a bit. You now have the experience you need to help design a *great* system. * Engineering-especially architecture-has skill sets that you really do need to build. In some ways it seems like things move very fast (thing about how fast kubernetes moves...). You must remind yourself, just like data science/math/etc, you need to build a *solid* foundation and then the rest will fall into place. Consider databases as an example. How much time have you spent trying to understand linear algebra? An insane amount. If you spend half that time trying to understand databases you will likely have a skill set that will last your entire life. * R&D engineering takes all of the skills that you needed from DS and adds all of the skills you need from engineering. In other words, you want to basically be a Math/DS expert, as well as a CS expert. Gulp. **What are some rough ideas of what you can do?** * You need *prioritization*. You can't learn it all today. What do you need this week, year, etc. You need some way to prioritize this. For instance, since you have been here you haven't heard the concept of *tensors* used once. Now, that doesn't mean you shouldn't learn it, but should it be a *high priority* that takes up all of your time...? Likely the answer is no. * You need to reassess your north star. What are you trying to achieve long term? What can help get you there? And (this is likely one of the most important things here) what do you need/want to achieve *short-term* that also is in line with your *long-term* goals? This is the sweet spot. Find things that fall into this bucket and focus heavily. * How can you find your niche, i.e. your opinionated stance on interesting things? It seems like that relates to *information*, *surprise*, *probability theory*, and so on. That is where you have great experience from the DS side. Use that to avoid unnecessarily theorizing. * Bring in psychology/front lines experience based ideas. What surprises people? What gets them going? How to find that? * Realize that being unbelievably innovative on this team is incredibly hard. Everyone is so smart and talented, and you are likely at least 8 months away from a senior level. Soak up everything you can while also really trying to make sure you are on ### Problems **What is one fundamental problem that you are trying to solve?** There are so many things to learn and focus on, how do I prioritize and figure out the correct way to spend my time and the best areas to give my attention? For instance, the problems we face today require certain skills, the problems we face in several months require another set of skills, and the problems we will face in a year require and entirely different set of skills. None of these are "fast" to learn. So I need to learn those that I need presently, while at the same time learning those that I will need down the line. k ### Goals **Short term** It is incredibly important to keep in mind the goals for the Fabric team (see [here](https://www.notion.so/unsupervised/AI-Planning-Q2-2021-da674f7a70eb4fbcade0bc73d9f8c4bb)). Anything that you can learn and do that helps achieve these goals will also help with career advancement, respect in the company, and so on. **Medium Term** The big 1-2 year goal for me is to learn a few more big blocks of Mathematics and Computer science, those that seem the most pressing/useful to understand at present. **Long Term** The long term goal remains the same: Build up enough of a *rich*, *deep* background in technical problem solving and understanding, and be able to wield that skill set to ensure *autonomy* and *control* over your schedule, finances, etc. This likely needs the following: * Demonstrated ability to solve real world problems, both technical, math related, CS related, etc * Ability to learn new things quickly. Postulate: most of the new things that come out are based on a preexisting base of fields and concepts. If you learn this base, you can learn new things rather easily. Learning this base is obviously of paramount importance. This will allow you to learn what you may need in order to solve new problems for companies in a consulting role, etc. * Ensure that your deep, rich base of knowledge means you don't need to learn many new things rapidly all the time. In other words, you don't want to end up on the hamster wheel. ### My questions to answer #### Root problem Build a deep, rich background and experience set in technical problem solving that will enable you to have *autonomy*, *competence*, *flexibility* and *control* over your life and how you want to live it. Shift over time to the point where you are selling your abilities and skillset more than your *time*. Never allow yourself to be in a position where you need to sell your time (long hours). Key Results 1. Direct on the job experience and expertise in solving real world problems that require technical expertise (Path Forward: Unsupervised). 2. Broad background (wide and deep) in key technical areas (Path Forward: Unsupervised, personal curriculum) 3. Demonstrable portfolio of skill set, communication, ability, etc (Path Forward: Blog, over time) Notes: * This is meant to ensure you don't end up on a hamster wheel. * This does not mean you need to have a PhD level of mathematical ability. #### The Most Important Thing The most important thing is clear: > Find the key areas of overlap where KRs (1) and (2) above overlap and dedicate your time and effort to them. ### Things to continue/start * Meditate, daily reflection and planning. This is more important than frantically reading. Again, this isn't to say *stop reading*, but rather accept that you learning rate is going to be through the roof right now and you need to couple that with extra reflection and planning, not simply more input. * Wednesday night deep work session! This should be on something that you are purely interested in, not work related (e.g. tensors, some branch of math, writing a blog post, etc), but it should be *technical* and move you towards a long term *goal*. * Podcasts. Podcasts have proven to be a great way to get solid, new information and ideas, in an enjoyable and easy to process format, that is a bit easier at the end of a long day compared to reading. This could be a great way to keep that randomness in your world and not simply stay confined to a specific set of ideas. * Book reviews and idea synthesis. This is just another form of reflection, now on a set of books that you have really grown to enjoy and love. * Note organization. This is a very low level of cognitive effort, but 15 minutes here and there will add up in ensuring you don't feel scattered. Again, this falls in to the bucket of reflection. ### Things to cut out or reduce * Pressure to read a ton. You are going to be learning at an *absurd* pace for a while to come. Keep reading of course, but let the books be ones that you enjoy (history, finance, etc), and not ones that are incredibly hard and challenging. There will be plenty of time for that down the line. This is not a need right now by any means and can significantly reduce your level of exhaustion. * Instagram, linkedin * Constant slack and email checking. --- Date: 20210516 Links to: [Decision by Design MOC](Decision%20by%20Design%20MOC.md) References: * []()