There are no experts in AI. And there won’t be for some time … if ever. That’s what I think anyway. Which is great since that means I can learn a lot about AI and have a career in AI or related technologies online itself without really having to go through years of study, right? I will come back to this in a bit.
Recently I played my first ever face-to-face FIDE Rating chess tournament. It was a 9-rounds, Swiss format, classical chess tournament which means each player will play 9 matches. FIDE stands for Fédération Internationale des Échecs (French for International Federation of Chess). There were 206 players, 92% were rated players (they have a FIDE rating), most have learned or are learning chess in dedicated academies and know all the openings and strategies. Quite a few had titles like CM, ACM, AIM, WCM, AGM. A CM or a Candidate Master is a player who has a rating of at least 2200. A FIDE Master (FM) has a rating of at least 2300. And so on.
My goal was to get a FIDE rating, for which I had to get at least 1 point. Probably sounds doable, but I am an unrated, untutored player, who plays only the Queen Pawn Opening and Queen’s Gambit and only quick formats (1 minute, 3-minute, 5-minute games). I have a 50% win ratio in the 1-minute Bullet format, having won 9187 games of the 18,541 I have played online.
But classical chess is different. It gives players a lot of time to think, plan strategy, and visualise possible outcomes of each move. You really can’t play for time, you have to play to out-think.
After losing the first match, I started watching videos on other openings and attack lines. There’s like tonnes of information for beginners and intermediate players and everything seemed doable. In the second match, I applied what I had learned … and lost again. Did the same for the next three rounds and finally realised that it is best I play the one opening I have been playing for the past many years. Learning takes times and the deeper the learning, the better the performance.
This is more or less how it is with using AI platforms.
Imagine you were given 10 days to create a staffing plan for a new coffee shop with details of how many people would be required to run the outlet, number of tasks, how many customers can be handled simultaneously, but you remembered today and deadline is tomorrow. AI can help you instantly create a 1000 page document or design a great presentation or generate a great video. But, if you had to explain the plan to professors giving reasons for your choices and reasons, it would become very difficult because the thinking and decisions were primarily taken by AI.
Now imagine that you are working at that very coffee shop, one of your colleagues has not showed up and one is on leave. You are alone with 7 customers in line and more walking in. Do you think AI or the plan you created can tell you what to do and how to handle those customers? If you have 1-3 years of experience working that job, you would know what to do without getting too hassled. For instance, I would come out of the counter, ask each customer for their order (noting on a paper or an app), and then request them to take their seats, letting them know I am working alone but will get their orders to them soon while they get on with their conversation or work. Then, I would make the orders, multi-tasking where I can and asking any new customers to take a seat and I would be with them shortly. There could be other approaches too but in customer service the most important thing is not to let customers get impatient or aggravated and that was bound to happen if I punched in orders, billed them, made them and then took the next customer. That’s what quick decision making is about, but decisions have to be correct and that comes from deep knowledge about time to make each menu item, time to bill, time to serve, etcetera, etcetera, etcetera.
AI is very similar. It requires deep knowledge for us to be able to use it affectively. Deep knowledge about subjects of our interest. If we are planning to study AI, we need to also know what we should study. (If you want to learn how to use an AI platform, just signup and use it – you don’t need to take courses on it. Really.)
But if you want to become a Data Scientist, then you would need knowledge and skills from different areas. After going round and round reading all the fancy definitions, I realised that Data Science is just analysis of information. That’s all that a data scientist does: analyses information. Why it becomes complicated is because we need:
- Programming knowledge related to Python and/or R
- Deeper understanding of machine learning (regression, supervised learning, etc)
- Good knowledge about statistical concepts and models
- Good data visualisation skills
- Domain Knowledge which refers to the area you have worked in or want to work in (for example, healthcare or food service or airport services or retail industries).
But where do you start?
One way of deciding what to study is to pick an industry you find interesting, find out the different roles (types of jobs) and how much those roles pay on an average, check what qualifications are required and then see what the entry requirements are for the university courses that offer programs.
So let’s take a look at what is called Median Salaries (average pay) in the Food Industry and AI-related roles in Food Tech. The following figures are from the Bureau of Labour Statistics, USA website) for 2024:
| Waiter/Server | USD 33,740 | |
| Cooks | USD 35,760 | |
| Supervisors | USD 44,140 | |
| Chefs | USD 60,990 | |
| Manager | USD 65,310 | |
| Sales | USD 66,780 | |
| Business Analyst | USD 101,190 | |
| Data Scientist | USD 112,590 |
Going by this approach, everyone would end up studying to become a Data Scientist. But that’s not what is happening, although that’s what all websites and education marketers are suggesting students should do – go AI on everything.
Another way is to pick a subject area you are interested in, check the job prospects are, and then proceed. Sticking with the example so far of Food and FoodTech, you could do a Bachelors in Food Science or Hotel Management and join the food industry and pick up required AI certification then. Or do a BSc in Data Science and then join the AI workforce in Food Tech and learn about the food business by research or certification as required. Another pathway is to do a BSc in Maths or Physics and then a Masters in ML/AI. Yet another pathway is to do a BSc Computer Science. What’s the difference?
Depth of knowledge at a practical level and a long term area of work.
A person who studies Mathematics at the undergrad level is likely to have a grounding in the fundamental mathematical concepts that underpin the development of ML and AI. A person who studies Programming at the undergrad level is likely to learn different programming languages (with some exploration of the underlying Maths) involved in the design and development of ML and AI. Either of these could be applied to any industry later – not just food. A person who studies hotel management will gain a practical grounding in business learning accounting, sales, food production, service, nutrition, food processing, purchasing, marketing, etc – it would be best suited for food industry roles but could also be applied to operations and business management in other customer-facing industries. AI and tech would be something you would evaluate as a purchaser to see how it benefits your company or industry.
By the end of the tournament, I lost 7 matches, won 1, and got a bye in one round, ranking 200 out of 206 players with 2 points, so I achieved my goal and will get a FIDE rating. The experience of my opponents was immaterial. What was relevant was that I stay focused on my goal, learn from every match and achieve my objective. A chess strategy is quite useless if I don’t know what my objective is.
Similarly, there is no point in studying AI to get a job. Study to learn and understand how the world works, how to build relationships, how to strike a balance between conforming and challenging norms and so on. Every graduate course will teach you how the world works. If you are not sure about what you want to study, a great way to find out is by taking short courses on platforms like GoSeeko (curriculum-aligned) or Deep Learning (AI-focused). A short course helps you test your interest and resilience (aim should be to finish the course). I take at least 2 short courses every month to figure out new directions. Once I know, there is more than just interest, I try to get an internship or a project or assignment in that area.
Or I play a tournament. Btw, I am a pro at several online games. But that’s a different story for another day.
Study AI with Goseeko: https://www.goseeko.com/landing/international-students/
Students from UAE, Register here: https://www.goseeko.com/landing/middle-east/