Beyond the AI hype: Inviqa on the shifting hype cycle

Ankit: So, when we talk about generative AI and innovative technologies in general, what does it mean for you personally and for your organisation?

Kaustav: It’s a great question. I’ve been an early adopter of technology from a very young age, always going after the latest breaking gadgets and gizmos and technology, and I think that permeates all the way through to my professional career today as well. But I think when it comes to innovation, for businesses like Inviqa and other companies out there in the world, the pace of change is just rapidly increasing at breakneck speed. And so innovation, I think, needs to become part of the DNA of almost every company that’s out there.

Ankit: I agree, and it’s in the DNA of Gate One. We have something called the Gate One Incubator, which is all about innovation, and every Gate One employee can be part of that incubator service.

Kaustav: I think that the whole premise around an incubator is amazing, because what you’re really doing there is you’re going out and you’re engaging with your clients and speaking to them about what their problems, their pain points are, where the gaps and opportunities lie. And I think that’s the right way to think about technology. And that’s where concepts such as an innovation lab can really start to augment what you do. Innovation labs in the past have always suffered from the age old problem of a bunch of nerds trying to come up with solutions, looking for a problem, whereas with the work that you’re doing in advance of building anything is truly understanding where the opportunities lie. That’s an opener to rapid development around prototyping and proof of concepting. And that’s where we can really start to bring those two sides together.

Ankit: And when we talk about these new technologies, we always follow them through a journey of a hype cycle. The five stages of hype cycle. From your experience, how is the hype cycle changing and what does it mean for generative AI?

Kaustav: I think the classic notion of the hype cycle is quite systematic, but in reality, it doesn’t always follow that pattern. When you look at a lot of organisations that are good at innovation and adopting new technologies, it’s the ones that really try to break the boundaries and in fact, try and do quite the opposite of what those technologies were meant to do. They stretch the boundaries of possibility to the limit and they create all of those edge cases that often break that technology. And I think that’s really interesting and fascinating.

Some of the technologies, whilst they might have been ahead of their time, I think back on technology examples like Google Glass when it first came out and then the Oculus Rift and the Magic Leap headsets and quite often they start off as technologies that are designed for the average consumer in their home. But as we’ve seen with a lot of these technologies, they’re not necessarily very successful in the consumer setting. But when you flip the coin and you apply those same technologies, like those augmented reality headsets, to an enterprise setting, you then start to find some really interesting use cases and I think things like voice augmented reality, VR and to a certain extent AI and generative AI has a lot of applicability in the enterprise.

But I think generative AI has totally changed that equation and really started to democratise how just ordinary people out there in the world are now starting to interact with artificial intelligence.

Ankit: It’s evolving at such a rapid pace it’s difficult to keep track of it. What advice would you give to organisations to avoid falling for the hype that’s coming with generative AI in social media?

Kaustav: That’s a great question. I think one of the really important aspects of any technology, let alone generative AI, but especially with generative AI, I would say is actually getting your hands dirty. There is no substitute out there for actually adopting these solutions, whether it’s Chat GPT or Midjourney or Stability. Being able to actually just go to the website, start prompt engineering and figuring out if you change your grammar, if you change the word or the sentence structure, if you rephrase your question. How does that affect the response that you get from these generative AI solutions?

And I think the other thing that people have started to get used to now quite quickly is that these solutions aren’t just question and answer solutions or something that you use to carry out a one off task creating an image. But it’s a two way conversation now with the AI it’s becoming a much more natural interaction. Whereas years ago you’d go to an e-commerce website and you would come across one of those bots that would ask you can I help you? Do you need anything? And you very rapidly figure out that all you’re talking to is a decision tree. It’s if this, then that type of conversation and the bot that you’re talking to doesn’t really understand what you’re saying.

Now with generative AI, whether AI is truly understanding what you’re saying is questionable. But the manner in which it responds to us now in a much more human like, natural, interactive way is what’s really the game changer.

And I’ve been seeing some amazing examples of generative AI over the last couple of weeks and months. The ability not only to converse in a chatbot like environment, but for creatives, for example, being able to augment their workflow, taking a still image and animating portions of that image and being able to produce video, audio and all manner of things that have started to become a copilot for them in their daily workflow and helping them enhance their capabilities.

Ankit: So once organisations have started on this journey, what are the next steps they have to take to really scale it in their organisations?

Kaustav: Capacity building is a really important place to start. And by capacity building, what I mean is companies really need to start to empower their employees to experiment. But not only that, but they need to set their employees up for success by providing the right environment, expectations and the learning materials as well. So access to courses, socialising and sharing the latest breaking training material that’s out there and there’s a lot of really great material out there that’s coming out week on week from various different experts in the industry. And to be able to systemise that, to identify willing individuals within a team, across disciplinary team, to come together and learn together, I think is part of what companies need to do much more and much faster so that they can figure out how to leverage this emerging space of generative AI in their day to day workflows.

Ankit: So keeping people / humans at the heart of it to drive innovation?

Kaustav: Exactly. And I think the more you start to dive into it and unpack it, the more complex it gets. I think learning and development for a lot of people can become second nature as part of their day to day work. But then when it comes to AI, and generative AI especially, I think the whole area of ethics as well is so important. Ethics and also making sure you don’t blindly believe the output from these technologies. I think a classic example that we’ve all heard of is how large language models hallucinate and make up facts and we still see that today, although it’s getting better and better and being refined as the technologies improve. But I think that’s a really important point to be aware of.

And on the point of ethics, I think that’s such an important area where these technologies are. At the end of the day, people have provided the data on which these generative AI solutions have been designed on and trained on and our biases naturally come into that data set. So when businesses are testing and adopting these technologies, before they unleash it into the wild, it’s really imperative that they test these technologies out. Look at the output, look at what they do with the images. So if your prompt is to get an image generator to create the likeness of a leader of a business, is it always producing an image of a middle aged male from North America? Or
is it truly representing a 50/50 split between male and female representation? The same can be said about ethnicity and background, language and so on and so forth. So I think testing and learning rapidly is really imperative.

Ankit: Thank you for your time today. It was lovely chatting with you and getting to hear your thoughts.

Kaustav: It’s been a pleasure speaking to you Ankit. Thank you.