breakthrough in Artificial Intelligence (AI) is not only the latest sensation to sweep the tech community, it is also crossing over into the mainstream media. Within just two months of its launch, the wildly popular AI chatbot, ChatGPT, has hit 100-million users, according to a new study published by investment bank UBS. That makes it the fastest-growing app in history. While schools and colleges are calling for it to be banned over fears of cheating and misinformation, Microsoft has recently announced it will invest a further $10 billion into its development.

Created by technology firm OpenAI, GPT – which stands for “generative pre-trained transformer” – is a state-of-the-art language processing AI model. Its ability to answer questions and interact in a conversational way with humanlike fluidity is capturing the imagination of both users and creators. But when it comes to this type of artificial general intelligence (AGI), the revolution has only just begun. We have not even scratched the surface of what is possible.

An AI-generated painting of the Vallée de Joux. The images in this article were created using artificial intelligence.

David Sadigh, Founder and CEO of DLG, has been a digital evangelist for over twenty years. After briefly dabbling in Computer Science in the late 1990s, he developed a specific interest in algorithmics and says he has long been studying Amazon, Netflix, Spotify, and other companies that use what are known as “collaborative filtering” algorithms to make predictive recommendations and increase sales.

David Sadigh, Founder and CEO of DLG (Digital Luxury Group)

Sadigh co-founded one of the first European digital marketing agencies (IC-Agency) that once built machine learning (ML) software to turn data into insights and actionable decisions for clients. He was also the mastermind behind the World Watch Report, which used ML algorithms to provide leading market research for the watchmaking industry for over a decade. In 2011, he founded DLG, which has since grown into a 130-strong marketing agency with offices in Geneva, New York, and Shanghai serving as a strategic partner for luxury brands.

Europa Star spoke with Sadigh about his views on the potential impact of AI. In this edited and condensed version of the interview, he discusses the technology’s wide-reaching implications both inside and outside the watchmaking industry.

Europa Star: Artificial Intelligence has been around for decades as a field of Computer Science. It’s everywhere. It’s behind everything from Siri, and TikTok, to self-driving vehicles. And suddenly, in the past few months, it seems to have moved into this new phase and become a buzzword, including in the luxury space. Why do you think it is?

David Sadigh: It is certainly not an overnight phenomenon. And it’s really important to highlight the fact that machine learning and AI have been around for years. I think what is happening now is the democratisation of AI. As soon as you have a computer and an Internet connection, you are most likely exposed to AI-based services such as content personalisation or ad targeting, often without even noticing it. What has changed is that now you, and the end-user, can access AI-based services and use them directly. What we are going to witness is the beginning of a major radical shift, where AI becomes mainstream everywhere.

“I think what is happening now is the democratisation of AI. As soon as you have a computer and an Internet connection, you can access AI-based services and use them directly.”

Why now? What is driving the breakthrough?

What is happening is that we are at the crossroads of many, very interesting paradigms that are creating an unprecedented level of change. And there are a couple of reasons for that.

The first element is, compared to when I started 20 years ago, the level of power that you can derive from a computer is much higher. Even the phone we have in our hands, in terms of capacity, is much more powerful than the computers we used to send people to the Moon. This exponential growth in terms of CPU (Central Processing Unit – the brain of a computer that executes commands) is one of the key elements.

The second element is the type of algorithm and all the neural networks that have been created to process information at scale. They enable links between information that was extremely difficult to decipher or to identify back in the day, resulting in more relevant algorithms.

Last but not least, as the cost of energy declines, it will provide more power and more storage of information in the Cloud at a cheaper price.

So, I think 2023 is not only the year of AI. We are entering into an era where ordinary people can basically develop their own abilities by using certain B2C- (business-to-consumer) accessible tools such as ChatGPT. And this is going to have a major impact. I think we are at the beginning of something that is probably as important, if not more important, than the early days of the Internet or e-commerce.

“We are at the beginning of something that is probably as important, if not more important, than the early days of the Internet or e-commerce.”

As this technology hits the mainstream, what does it mean for business, and how will that translate to our day to day?

I think that many jobs will be impacted with the new tools that we are seeing, especially jobs where the cost of labour is high and jobs that can be replaced or partly replaced by the computer. For example, I tried to use ChatGPT to help me craft some clauses for a commercial contract and the result was outstanding. I showed it to a friend of mine, who was a lawyer and it made sense to him. So, obviously it doesn’t mean that ChatGPT will replace lawyers tomorrow. But it means that, probably in 80% of cases, some basic operation can be replaced by ChatGPT or by a new version. So, I think we are witnessing the crossroads and we will be seeing more new use cases.

How would you advise newcomers to this technology to navigate the new landscape?

My first advice is to really spend time understanding it. That doesn’t necessarily mean going in-depth and trying to manage the nitty-gritty of AI. Obviously, that will require some other type of knowledge and mathematical, probably algorithmic, understanding. But try to understand the main principles and how they could be applied to your business.

And for most executives, I think it’s becoming urgent and critical to take some basic introduction classes, also to play with some of the tools online on their own. But don’t just think: ‘Oh, this is a new type of trend like NFTs or blockchain’. Because I think, in fact, this is something much bigger.

“Don’t just think: ‘Oh, this is a new type of trend like NFTs or blockchain’. Because I think, in fact, this is something much bigger. ”

And how much bigger should they expect it to become? What sort of AI applications could they see?

There will be a lot of different applications and use cases across the board. Let’s take a company and start with sales and marketing, say, market research. Right now, you have people compiling lots of data into Excel sheets and creating different types of charts. At DLG, we did a couple of tests. And we managed to get access to maybe 60-70% of the raw data that was needed to perform some kind of market research. Of course, there is risk involved in terms of the quality of the data. Is the data quality reliable enough? But this is just to give an example of market research, and how it is going to be disrupted by the new tool.

Now, if you go deeper, say, pricing strategies you have seen on e-commerce websites, when you visit an e-commerce website, you have what we call dynamic pricing. It’s like when you buy from an airline, the price will change depending on the user, the context, and their behaviour. One should expect this to become used much more with many more companies adopting dynamic pricing strategies. So, these are just two examples of how machine learning will help create new ways for companies without needing as much human intervention as before.

What about the watch industry specifically? How could AI be applied in your view?

I could share with you many examples about the watch industry. Once again, across the board, I don’t see many things that won’t be impacted. Right now, if you look at the creative process for a watch, we are already seeing the first use cases of artificial intelligence tools which could help in designing, or in the ideation process of crafting new types of watches. So, I think we will see the first watches coming to fruition which have been either fully or partly designed with AI in a couple of months. You will see that developing very quickly.

Furthermore, the supply chain is going to be disrupted. There are a lot of elements coming from AI that will help automating operations, such as price arbitrage when buying a specific type of material, and so on. I can also think of the warranty tied to different types of services, either being created or technically priced. So, when you buy a watch, it will tell you based on your profile that this is the type of guarantee extension the company can offer you.

Then, we have the customer service department, sharing all the exchange of email they have with you, for example, and being able to craft automated responses. So, those chatbots that you see now will become much more autonomous and able to enter a conversation, where you may ask, “When is my watch going to be delivered?” Not only will the chatbot give you the FedEx ID so that you can track your watch, but it may also ask if you have a preferred time in the day for delivery, or if you want to change something, and then it will handle the operations on your behalf. A bit like if you would be conversing with a highly-skilled and professional customer service specialist, in fact. So, there are many different elements that are going to happen, and it is very interesting to witness this new revolution.

“We will see the first watches coming to fruition which have been either fully or partly designed with AI. You will see that developing very quickly.”

Given that you once ran a business with a background in ML algorithms, will you be positioning DLG to support businesses interested in this area?

Yes. The first element we have started doing with our clients is workshops, having discussions and showing them some use cases to help them think and transform their business. Some of them realise already. I think some will realise in the weeks or months to come, to make sure top managers, middle managers, and people in charge of making decisions with company strategy are informed of the changes, and the risks, that are going to happen.

Then, we will probably do much more. We already have some tools that we use, some software to personalise, for example, the client experience. These third-party solutions, such as Insider (a platform for individualised, cross-channel customer experiences) or Contentsquare (digital experience analytics) are also based on ML algorithms. They allow our clients to create an elevated type of customer experience, and to improve conversion rates on their e-commerce websites. And we will invest in the field because we do believe that the more our clients can grasp those topics, the more they can leverage the technology, the more competitive they can become.

It is fair that a lack of knowledge or understanding may cause fear. For years, the likes of Elon Musk and the late Stephen Hawking have cautioned against AI as “our biggest existential threat”. There was an open letter back in 2015 which they co-signed (with experts in the field) to sound the alarm for the risks of AI. So, how has that changed? How is it less scary to people or threatening to humanity, do you think?

First of all, I agree with them on the fact that it is scary, and it needs to be regulated. Think of all the major radical technology innovations, such as the radio, the television, the Internet, mobile phones, which have been launched in the last century. All of them have required some kind of regulation. AI will be no exception to that. And we will need to have very strong regulation and extremely clear rules, policies, and guidelines, so that software developers and all the key stakeholders involved within the machine learning/AI industry will have to follow and observe those rules.

On one hand, we are already seeing big biases linked to those technologies. On the other hand, we are also seeing right now venture capital companies investing into AI, like a type of AI police, to be in charge of controlling all the AI so that there are not too many biases in the way the AI operates. And on that point, I’ve played with ChatGPT, which is doing, globally speaking, quite a decent job at not falling into the trap of racism or discrimination. But it’s far from being perfect.

So, I think for the whole new ecosystem that is being created, the regulatory part will be an extremely critical and decisive component. It needs to ensure the long-term sustainable development of AI is aligned with the ethical and the moral standards that we want to see within our societies.

“All the major radical technology innovations, such as the radio, the television, the Internet, mobile phones, have required some kind of regulation. AI will be no exception to that.”

Where do you see AI going in the future?

The future of AI is difficult to predict with certainty, but it is likely that we will see continued advancements in AI technology, including improved natural language processing, better computer vision, and more sophisticated machine learning algorithms. AI is likely to become increasingly integrated into various industries and domains, transforming how we live, work, and interact with technology. However, ethical and societal considerations will also play an important role in shaping the development and adoption of AI in the coming years.

And by the way, this last answer is provided by ChatGPT.

To conclude, and as Alan Kay, the “father of personal computers”, once said: “The best way to predict the future is to invent it.”