🐘 Bonus Short: Touching the Elephant—Our AI Journey with Atif Khan

The magic of AI is that it no longer requires learning a new language, like how to program. AI moves computers from data machines to cognitive companions. The best part? There's never been a better time to start using it.

Episode Transcript

(00:00):
There's a story that I grew up listening where people were taken to a dark room and they had never seen an elephant before.

(00:07):
And they were all allowed to touch different parts of the elephant.
And then when they came out again, everybody described the elephant differently.
Because the reality sets in based on what your experience is with AI, what you've experimented with it.
So the beast looks different to different people.
And this specific beast is not just an elephant, it's a dinosaur.

(00:28):
It's big enough to cause all kinds of confusion as well, right?
So I think not learning it, not understanding it, I think the journey literally starts from there.
If you don't go back and start that journey, you'll always be perplexed, always be confused in terms of like, what is this thing?
What can it do? Am I using it the right way?

(00:49):
Am I riding an elephant? Am I riding a dinosaur? What's going on here, right?
And the demystification of the technology from, is going to take my job away to magic or to, it's just another tool that I need to figure out how to apply it, how to use it.
I think that journey, unfortunately, we all have to embark on our own and figure things out on our own.
We all have to go to the dark room and touch the elephant and then perhaps go in that dark room multiple times to understand what is the bigger picture here?

(01:17):
What's the bigger animal here, right?
Yeah, I actually just got asked this morning, how do I keep up to date with industry trends?
And I said, keep up to date with AI, read the highlights of the news, things that are coming out, but then go and play with it.
Yes.
There's no other way to bridge the gap other than go and touch the elephant.
Yeah. People learn differently, but I think not getting your hands dirty with it is a mistake, right?

(01:41):
You have to play with it.
Yeah.
It doesn't matter in which capacity and then, interestingly enough, you're not, there's no barriers now, right?
You're not, like, you don't need to learn a programming language to play with it.
You don't need to learn by a specific hardware to play with it.
Like, you can literally just get on a browser, talk to it and get a feel for what it can do.
Now, maybe you're not building the next solution with it, but at least you're understanding what it's capable of.

(02:06):
In my mind, there's no reason why you should not be touching it today, right?
And why should not be dreaming bigger, dreaming differently with it today?
But I also understand it's a journey, right?
Not everybody's going to be switched day one or day two.
It's like, oh, now I'm an AI guy and I understand everything.
It's not that matrix thing. It's a journey for everybody.
It's a different journey for everybody. It's not even the same journey for everybody.

(02:28):
Honestly, I consider myself a bit of a leadite.
I'm a lagger when it comes to adopting new technology.
And when I see something like chat GPT and all the hype, it feels like it almost makes it less interesting to get involved with
because people are saying, oh, no, the robot can talk now.
So it's kind of a bunch of all our nukes.

(02:49):
And it's not realistic enough to engage with.
And I feel like that might not be a bad thing.
There's never a late time to get involved.
It's not you need to be an early adopter or you're going to miss out.
It's just get involved.
It is changing pretty fast. It's now almost two and a half years.
So was it I think October, November?

(03:12):
It was somewhere around that time, two years.
I think October.
Then we figure out, OK, so chat GPT is now in the hands of everybody.
It was text only. It was still producing.
Some people were still like, oh, I'm not so sure.
Some people are believers. Right.
My own journey was I'm going to have to try different things.
And so it writes well. But for example, it had no notion of counting.

(03:33):
It had like it had problems even back then. Right.
Yes. If you just look at that, that two and a half years of where we are today, it's mind blowing.
Yes. And the question is, do I learn it now?
And don't worry about it having missed out on not learning it two years ago.
Or I think the answer is it's a continuous journey.
There's no right time to jump into it.

(03:56):
Every time is the right time to jump into it.
It's going to keep changing.
Don't learn just the technology and learn the fundamentals behind it as well.
You don't have to learn the technical part, but understand how it works.
It's like getting to know a new person.
And that person is also rapidly growing and changing and having its own thoughts as well.
Right. It's like it's just a from that perspective.
What would you say is the biggest difference in how you use it between then and now?

(04:22):
When it started out, we were only looking at from generated for purposes. Right.
So even a lot of pre-GPD 3 work that we did in our lab, it was mostly around generating synthetic data for testing and things like that.
The idea wasn't in my mind, wasn't there that it's going to be within a few years at a point where it's a cognitive companion.
Right. That I can talk to it, I can show it my ideas and I can have it polished them.

(04:46):
I can have it reflect on them.
I can literally, the way I would take my ideas to a co-worker and says, hey, can you, what do you think of this thing?
And we all do that in our day to day lives at work, reflect things with, talk things out with other colleagues.
I had not imagined that it's going to be in two years, it's going to be at a point where I'll be able to just engage with it in a meaningful way.

(05:07):
Besides this generating synthetic data. Right.
The code part was exciting, but code code part for me was just a natural evolution of text generation.
Because code is unlike natural language, code is so well defined by its the sess, the good.
It needs to be parsed. It needs to be able to compile.
So there's well defined tools.
I wasn't surprised when they said, well, we looked at entire GitHub and then learned how to code.

(05:32):
That was understandable, made sense.
Natural language was always been one of those things like, okay, so it's hard to get into it.
It's not crisp. It's not well defined.
People can mean different things or different languages.
So the text generation part was interesting.
But this idea that it can actually reflect on what I'm saying.
And it's almost like taking a whole bunch of experts and mimicking them into a bottle and say, or into a system and says, here you go.

(06:00):
So instead of just talking to one expert or talk to one colleague, I can potentially tap into the voice of many experts and many colleagues from different perspectives as well.
It's just a matter of how do I engage with it? Do I engage with it as a colleague? Do I engage with it as an expert?
And that opened up a lot of possibility in terms of how you use these things.
How do you even? It's not just about product anymore.

(06:22):
It's about personal growth. It's about learning. It's about consuming content.
It's about the speed at which I can consume content and change content.
Everything changed for me.
Yeah, I really like the term you use, which is cognitive companion.
Oh, nice.
That's beautiful.
And I feel like it's a new category that I don't know anybody predicted two years ago.

(06:44):
But it's actually said it's not text processing. It's closer to thought processing.
Yes.
And I've talked to it as Socrates before, for example, I hear people talk to it as Steve Jobs.
If you want that kind of opinion and get bottle the expert and then use them as a cognitive companion, it's awesome.
I started this podcast because I wanted to stand at the gate of businesses using AI and see what separated hype from lasting impact.

(07:12):
When the cities had walls, you had to go into the city to do business at the market.
So if you wanted to talk to someone, you waited by the gate until they came in or came out.
Do that enough times and you could talk to everyone.
That's what I want to do.
Stand at the gate of people doing business with AI and talk to them.
See what they do and why they do it.

(07:34):
If you know someone that's making an impact in the world of AI, would you connect them with me?
Find me on LinkedIn or shoot me an email at [e-mail].
That's [e-mail].
Thanks for listening.