HN Explorer

Ask HN: How to Become an AI Consultant?

11 points by jdthedisciple ↗ HN
It looks like being even a talented run-of-the-mill Software Engineer with a good CS foundation is not gonna be enough to make real break-throughs career-wise in the coming times.

So I was wondering if you are just that, what's the best way to transition into becoming a full on AI expert and consultant?

Is it best to be primarily a daily AI user yourself? Is it best to be an AI developer? Is it best to be an AI researcher?

And how do you even get started with that transition. There is an exhausting amount of things going on and coming out on a daily basis!

10 comments

[ 3.4 ms ] story [ 31.0 ms ] thread
rmaiti
AI is ONE of many CS disciplines, and if you want to specifically concentrate on it then sure, in the coming years, you may be able to reap some degree of alpha due to this. But in my opinion, the focus on AI is going to plateau, and everyone, regardless of their disciplines, would be using some sort of generative AI tech. in their work. Commonplace application of AI would benefit the early movers, but in the end, it likely would not result in the wholesale replacement/eradication of programmers IMO
jdthedisciple
Ah I'm not talking about replacement of usual programmers at all.

More about what's going to be an opportunity for huge leaps for a while.

Since MILLIONs of businesses are most likely asking one question right now:

"How can this AI stuff make ME more money!?"

Which is where a new breed of SE+AI experts will come in.

Question is: What's the straightest way to become just that.

rmaiti
Ah, I see. Something that I hear people talk about is "Prompt Engineers", but the way current research is going, this is something that is going to be automated, or instruction-tuned LLMs would make it obsolete.

Rather, you can focus more on learning LLMs in-depth, finetuning them for a specific niche application, learning how to parameterize them appropriately, etc.

ipaddr
You've already missed the boat and others have a big headstart. I would focus on the next big thing.

People have already figured out AI won't replace the developer in todays development process. At some point it will produce programs that work.. but programmers will be needed to connect those to other programs. Once AI designs the entire system at that point you will not be needed.

You have boring skills you can consult with now.

aethelyon
It depends on what consulting you are doing. You need to match the experience/expertise to the services. Most organizations need help on strategy, implementation, and hiring.
jdthedisciple
Sure but that's the question:

What consulting should one be doing, if the overall goal is to just "use AI to improve business results"?

And how to get there.

tikkun
It's kind of a chicken and egg situation. The biggest thing businesses need is someone that they trust who has gotten great results for 5-6 of their peer businesses, such that they get recommended to use this person/org and that when they talk to this person/org they know their problems so well and can talk confidently about what the business is dealing with, will need, and can attain.

So, you need to pick the thing where you can get great results for 5-6 businesses and have them as case studies - most likely no one will ask to see the case studies, but it's more about having the internal skill you get from doing them.

Some things that might help you figure that out:

1) Which type of businesses and buyers (i.e. titles/roles/problems) would you pick if you were picking the one that'd make it easiest for you to reach your first 10 customers? For example where do you have practice in selling to a certain type of org or a personal network that includes many of a certain type of buyer in your first degree network

2) Which type of business and buyers do you believe you'll be able to get the best results for?

3) Which type of buyers do you like the most and feel like you can best teach, and where they'll want to listen to you teach them? Businesses (and people in general) like buying from people who have successfully educated them about things that they care about

If there's no obvious answer to the above questions for you, then my suggestion would be to brainstorm your best options, and go try them all out either in parallel or in sequence, and then if one of them seems to work well for you, double down on that one.

paddw
You have to be a good talker. If you are not you will struggle.

Find a friend who is willing to pay you, or consult for low-pay/free to build up a client portfolio.

Then try to get referrals. The hard part is not understanding new AI stuff, but building the consulting business.

muzani
Propose solutions that save money. How does AI save money? If I knew, I wouldn't hire a consultant.

Lots of people are also doing this thing where they go, "if you pay me, I'll give you the details." No, you have to prove your competence long before that. Hang around LinkedIn or write blogs or something.

comprev
Ask ChatGPT. Surely they would know how!