4 ms·
Well first of all I’d say that your goal shouldn’t be to stay in the same role, even if it was keeping pace with inflation (in future it likely will anyway, whe
by Tycho 3y ago
Well first of all I’d say that your goal shouldn’t be to stay in the same role, even if it was keeping pace with inflation (in future it likely will anyway, when inflation returns to normal levels). By evolving or changing your role you can double or triple your income over 5-10 years. I remember my first two full years with a decent job I got 20% raises, then was disappointed to get 2% the third year, boss told me I should expect that from now on. So that would mean 35 years to double my then salary… in reality, I changed role and more than doubled it within a year.
If you’re 25, you might as well spend your evenings trying to upskill. Make time for important things, but trim unnecessary leisure and invest in your skills/expertise/network. It’s much harder to do this when older.
Personally I continued to improve my technical skills, especially in getting into new areas like AI, while also building considerable domain knowledge, and taking on management responsibilities. You don’t really have to compete directly with yon 10x STEM rockstars - go for breadth, not proficiency. Someone who understands a lot of areas fairly well, and can communicate and document well, is incredibly useful to an organisation.
- helboi4 3y agoDid you have to brush up on maths to get into AI properly? Theres a short course at a STEM university in AI that an old colleague of mine did. I kept considering it, but it says it requires strong maths skills. I dropped out of maths at 17. I don't think I'm so stupid I couldn't get past a 17yo level of maths, but I do think that I'm severely behind and it would take a lot to make up for that.
- Tycho 3y agoI didn’t have much formal maths beyond 17, just a year of university statistics. I think it paid to spend some time studying linear algebra and statistics (reading Taleb’s books helped make it a more interesting journey), learning to read mathematical notation, and just getting kind of comfortable with the field so that you know when to ask an expert, but you don’t need to go very deep. Most AI professionals probably spend most of their time on data wrangling and engineering, not using any advanced maths. And then in order for their teams to achieve anything useful they need to work with “business” people who have domain knowledge. Also most ML algorithms I’d say can be grasped conceptually without being a big maths person, and then you have a sense of where to apply them. But over the years, you’ll pick up a lot of stuff anyway. Plus, with modern LLMs, it’s anybody’s game frankly. There might be some really tough maths on that course, but I’d argue it’s not exactly necessary. There’s undoubtedly a lot of people doing high paying “AI” related jobs whose main qualification is being able to identify a use case and coordinate a project staffed by engineers and data scientists.
- helboi4 3y agoThanks for answering my questions. Could you explain what you mean by Taleb's books. I don't know who that is and what books you are referring to. I tried to google Taleb but there are a lot of people called that.
- Tycho 3y agoNicholas Nassim Taleb, author of Anti-Fragile and The Black Swan. I recommend those ones. He writes about the role of uncertainty in life. Certainly less dry than a statistics handbook, and drives home a lot of important concepts.