SDS 424: A Symbiotic Relationship With AI

(00:14):
Hey everybody, super excited to have you back here on the show. Today we’re going to be talking about something interesting, about learning from artificial intelligence. So a few episodes ago, on episode 417, we had Art Shectman, and we spoke about data engineering and product development. Well, Art later followed up with me with a blog on the BCG website. So BCG is the Boston Consulting Group, and they recently did a very interesting survey of 3000 executives worldwide asking about artificial intelligence, and there’s some interesting findings which I’d like to share with you today. 

(00:54):
Who will this be useful for? Well, it will be useful for you if you’re an executive and you want to implement AI in your company for sure. Some very interesting ideas or insights on what steps are required in order to see the maximum return on investment. And also if you are a data science manager or an individual contributor in data science or an analyst, this will be useful to you as well in case you have these conversations with the upper management and you can contribute to them and point them to this report later on. 

(01:27):
Okay, so 3000 executives were asked questions, and the first insight here is that things are going good for AI. Well, the sentiment around AI is improving over the years, and the insight here is that in 2020, about 60% of the companies, based on the survey, had an AI strategy. Two years ago, in 2018, it was only 40%. So that has improved. However, the real tricky part, the real question is what is the return on investment? What kind of results are these companies seeing? Well, the insight here is that unfortunately, most companies aren’t seeing a significant return on investment. Only about 10% of companies report significant financial benefits from implementing AI. 

(02:26):
Hence the question, what’s going on? How come more companies have an AI strategy, but still only 10%, so 80% of companies have an AI strategy, but only 10% of companies actually are reporting significant financial benefits from implementing AI. What is going on? And this report provides some insights. We’ll link to it in the show notes, of course, or you can just Google for it. It’s called Are You Making the Most of Your Relationship With AI? It’s October 20th, 2020, so quite recent. So what are the insights? How are the companies that are doing well, how are they making the most of this relationship? 

(03:03):
Well, the report highlights that there are three main stages of how you can adopt artificial intelligence. The first stage is when you simply have the right data, technology, and talent in place, plus a corporate strategy. So if you have those four elements, then a company… Let’s look at an individual company. So we take a company. If that company has implemented those four steps or four elements, right data, technology, talent, and strategy, they have about a 21% chance of getting a substantial financial return on investment. So they have a 21% chance of being in that 10% of companies that are getting value out of AI. 

(03:53):
Now, what’s the next step? Well, the next step is if you add to that equation if you add the ability to iterate with artificial intelligence. So basically they’re saying the ability to iterate on AI solutions with business users. The way I understand it is that you don’t just deploy a model, or you iterated through its results. Yes, when you’re building it, but also once you deploy it, you then interact with business users, see how they interact with it. So you get some feedback for the AI from the people using it, and you incorporate that feedback in the model or in your artificial intelligence solution, whatever it is, then your company’s chance of seeing a substantial financial returns on investment goes from 21% to 39%. So almost doubles, simply by including this feedback for the AI from the business users. 

(04:51):
And finally, the third step in adding AI to your organization is the real key to success, because it takes the percentage from 39 to 73%. Quite high. Three chances out of four of seeing substantial financial returns. So what is this third step? Well this third step is the ability to learn as an organization, and they more describe it here as by bringing together human brains and the logic of machines. So what does that mean? Well, that means that not only does now AI learn from human feedback, humans also need to learn from AI feedback. 

(05:44):
And this is across the organization as a whole, meaning that systems will have to change, processes will have to change, architecture of IT infrastructure will have to change. The mentality of humans operating alongside AI will have to change. A lot of things will have to change. It’s a lot of work. But it’s worth it because that means that percentage of getting substantial financial return goes from 39% to 73%. So a huge increase. 

(06:17):
They go more into discussing the different types of symbiotic relationships. I’ll leave it up to you if you’d like to explore this article further. Again, we’ll link to it in the show notes, or you can find it online. But what’s interesting is another insight that they provide is that to support this idea that a lot of things need to change to incorporate AI. You can’t just put AI into one process and hope for a huge financial return. That maybe works for very few companies. As we discussed, the chance of that having a success is either 21% or 39%, depending on how you deal with the feedback from human to AI. 

(06:53):
But what they found is that if a company incorporates artificial intelligence, not just in… Make changes, basically if they make changes to processes to better work with artificial intelligence, not just in one, not in two, not in three, but in multiple processes or systems in the business, companies that really go full in, all in, and go all the way and do as much as they can to incorporate AI, they’re five times more likely to realize significant financial benefits than companies that only do one change. That’s a big insight. 

(07:32):
So basically the takeaway here is that AI is likely, is very likely to fail if we just put it into one process and like, “All right, let’s automate our call center, or let’s automate this predictive maintenance part of the business,” or something like that. It’s very likely that you won’t see significant financial returns. The key to seeing significant financial terms is adjusting many processes, many systems, getting many people on board, creating those feedback loops from AI to humans, from human to AI, from AI to itself, and setting up the system in such a way that the whole organization is undergoing a change and is recognizing that, “Oh, okay. So now we’re using AI and we want to get this financial return. We want to get this benefit.” It’s a lot of work indeed, but those companies that do that work reap the financial rewards. 

(08:29):
So there you go. Hope that was useful. You can find this on the BCG or Boston Consulting Group website. It’s called Are You Making the Most of Your Relationship With AI? And send this article or this podcast to somebody who you know who’s maybe a manager or executive. I’m sure they will appreciate it. Boston Consulting Group is one of the top three management consulting firms, and something coming from them in partnership with MIT Sloan Management Review has definitely got a lot of weight. On that note, I look forward to you back here next time. Thank you. And until then, happy analyzing.