[Discussion] How can we do the transition to AI/ML technology domain?

First of all, I would like to thank everyone who participated in the discussion and spare their valuable time, and shared their knowledge. 

We had a great discussion which we can say was fruitful. I will try to summarize the discussion.

Best way to learn AI/ML is to

1. Start working on a project - Hands-on experience!
2. Guidance is necessary, get a mentor!
3. Be surrounded by like-minded people.

Of course, this must be preceded by clarity on why you want to do the transition? Clarity can be of either of

• a) Building a career in a particular domain. For example, I (Dhawal Banker) am learning AI/ML for the Computer Vision domain because that’s my area of interest. OR
• b) Exploring the technology to get a better understanding before making an aware decision.

But in either case, there should be clear measurable goals/targets to achieve. 

Coming back to transition. Typically there are two approaches:

1. Structured learning via study programs
2. On Job learning

Structured learning

We can find many programs addressing the learning gaps for AI/ML from many great Universities and even from many individuals. One can easily find it by a quick Google search. What should we expect from such programs? 

• Covering basics and prerequisites for around half the duration
• Theory and significant time on hands-on projects. 
• Continues guidance, a support system to keep us in the correct direction
• Networking and learning with fellow students
• 6 to 12 months duration
• Support for getting a job

On job-learning

This is also very well structured learning from the employer's perspective. The environment for ideal learning is typically available in a company where

• You should be enthu for learning and you are included in the AI/ML team
• You are given a well-defined problem to work on 
• You are part of a team that is working together to help and guide 
• You may have to push yourself for learning to make a bit of structure for general skills, but very apt for problem or domain-specific learning.

3rd Approach

There is also a 3rd lesser discussed but tried approach is of self-pace learning. It’s an online learning era. We can find most of the learning information available online. Even some really good structured learning courses are available for free on Coursera, MIT OCW, edX, NTPEL, etc. But it has a different set of problems to overcome 

• There is a high chance of getting lost in the Rabbit hole. But one can also have total control over what to focus on and spend more time on. One needs to be really focused on the define measurable goals with the above-mentioned self-clarity of measurable goals.
• There is a definite lack of guidance. You surely need a mentor which can help you float through uncharted waters. 
• The journey is lonely. It’s typically difficult to walk through a somewhat difficult path alone. Here online communities and groups come into the picture. 

AI for Computer Vision is one such group. We typically have regular meets on Thursday evenings, where we network, find solutions to our problems for a long journey, discuss specific AI problems, etc. Basically, we try and help each other get the transition into the unknown world of AI for Computer Vision.

If you are on the same journey, we would recommend you join the group. You can either join a LinkedIn Page, discussion group, or Meetup page. 

You can also contact me to become part of our Whatsapp Group.

Thanks,

Dhawal Banker