Not a day goes by without a mention of ChatGPT or artificial intelligence. The AI buzz reached unprecedented heights around the new year, but is it just all hype, or are we witnessing a revolutionary leap forward? We had an interview with Antti Merilehto, who has spent the last eight years immersed in AI. We wanted to find out where exactly we stand with AI right now.
Could you introduce yourself?
I’m Antti Merilehto. In this professional context, I am an entrepreneur and an investor specialized in early-stage technology companies.
I train teams, executive groups, and occasionally individual leaders in companies for a living. Over the last eight years, I have specifically worked with artificial intelligence.
Today we are talking about artificial intelligence. Let’s start with a light question. When and how did you last use AI?
I used AI just eight minutes ago. I asked ChatGPT for some ideas for a webinar. I already had the content ready, but I wanted to tweak it to make it more sales-oriented. I asked for help from my AI colleague, who has a knack for rephrasing my ideas in a slightly different way.
AI is such an awesome workmate that I can leverage its capabilities anytime and anywhere.
In 2018, you published the book “Artificial Intelligence – Travel Guide for Leaders”, which was the best-selling business book of that year. What have been the biggest changes in the field of AI since then?
A lot has happened in the meantime—both visible and invisible changes. In business, artificial intelligence is now frequently used for making predictions. It helps, for example, in forecasting which customers to contact at any given time and where to allocate resources effectively.
The biggest change is that very few medium or large companies operate anymore without some advanced analytics spinning in the background.
Of course, we must also mention ChatGPT, which is a single language model interface. In the grand scheme of things, it doesn’t really matter what kind of technologies exist until people can actually get their hands on them. ChatGPT provided a tool that anyone can use very easily. Consequently, the interest in artificial intelligence has exploded dramatically.
For my business too, ChatGPT has been a white swan.
Artificial Intelligence has become an even bigger topic with ChatGPT. Is AI worth all the hype?
If you frame it like that, then probably not.
In the AI conversation, there are people involved who understand about as much as a pig does about windmills. Often, it’s not understood that the quality of data determines the outcome in almost every situation.
If three companies adopt the same AI application, the outcomes are likely to be very different. The toolset is identical, but the results depend on the quality of the available data.
In my opinion, AI isn’t worth all the hype it gets, but in some aspects, it’s an even bigger deal than the hype suggests. It already enables us to achieve many great things.
According to Forbes Advisor, 97% of business owners believe that ChatGPT will benefit their business. Is it really such a significant game changer?
Three percent of respondents are wrong.
I have an office in Helsinki’s Punavuori district. In the neighboring block, there’s a bakery run solo by a French baker. He bakes a certain amount of pastries each day and delivers them to nearby restaurants. He doesn’t need ChatGPT.
If even a single person in the company uses a keyboard, then ChatGPT is definitely beneficial. It’s an industry-independent tool that’s equally useful for solo entrepreneurs and global market leaders alike. With it, anyone can whip up an innovation voucher application in no time without needing help from a communications expert.
Recently, we were just thinking about doing some pilots, but now many companies are creating AI strategies. What exactly has changed?
Even academics who study algorithms and AI professionally wouldn’t have believed that language models could evolve this fast.
Language models like ChatGPT suddenly made AI a more comprehensible matter. This shift brought AI into the corner offices of businesses. In many companies, top executives had the revelation, “Holy smokes, this can enable a whole bunch of things.” There was also a wake-up call across businesses realizing that the rapid development of AI is going to change numerous ways we work.
I’ve also noticed that conversations with clients have been different from before. Attitudes have changed. Knowledge and understanding have also grown tremendously.
Artificial Intelligence is no longer seen as just a separate component that’s slapped onto a machine. So, in discussions, the focus isn’t on what AI could do but rather on identifying business challenges and figuring out how to solve them. Sometimes, AI is the solution to the challenge.
In a nutshell, AI is now being advanced not from a technology standpoint, but a business-driven one.
Large companies are naturally at the forefront, but what is the relationship between SMEs and AI? Can AI be leveraged even with smaller resources?
I’ve even been training solo entrepreneurs on how to use AI tools. What does a solo entrepreneur do? They communicate with their customers.
As a sole proprietor, one naturally tries to be the master of their own destiny, like Jari Parantainen, and package their own services and expertise. In this endeavor, artificial intelligence is an excellent assistant.
Imagine that you become a vinyl record dealer. With the help of artificial intelligence, you can get the establishment papers done in a flash, and when a potential landlord makes an offer on the rental space, ChatGPT can help you come up with reasons why the rent could be, say, 20 percent lower.
It’s particularly the small business owner who massively benefits from having a small army of (AI) coworkers at their disposal.
Therefore, leveraging artificial intelligence really isn’t about the resources available.
Do you have any examples from the SME sector?
There are tons of examples I could give.
Take, for example, the energy sector. Last winter, not a single company in the industry had enough customer service reps to handle customer inquiries. What if we had collected all the old feedback and data from customer service conversations and used them to train a language model to respond to customer questions?
Antti would still be there as a customer service rep, answering queries, but ChatGPT could handle a large volume of contacts and immediately respond to commonly asked questions.
Let’s introduce a completely different example. If you operate in an industry where the buyer is a politician, then you’re essentially doing lobbying work.
I’ve been helping numerous influencers leverage ChatGPT, as ministries and such are churning out thousands of reports that an influencer should, to some extent, be aware of. It’s a utopian notion to think that one person could get through all of them. With ChatGPT, sifting through and analyzing information becomes vastly more efficient than ever before.
Consider what this change means for all the lobbyists and lawyers!
What have been the most amazing AI applications that you have come across?
I’ve picked out one boring but sensible example, one funny, and one business-related example.
Boringly formal but immensely important, an example relates to slowing down climate change and how artificial intelligence can enhance energy efficiency. This example initially inspired me to write a book on AI.
Developed by Google, DeepMind was able to reduce the power consumption of data centers by 40 percent. As global energy consumption swells every couple of years and climate change continues to accelerate, how can we reduce energy usage in all existing entities, such as data centers? While this might sound like a rather invisible and even dull issue, it significantly impacts our ability to keep this planet habitable.
From the “Funny” category, I want to highlight internationally successful photographer Antti Karppinen. As a visual person and an amateur photographer myself, I have been utterly amazed by how Karppinen utilizes AI in his art. He recently released a series called Ridiculously Strange Finland, which challenges us to question whether the images could be real. Of course, they are not, but they look just like real photographs. If you consider the purpose of art, in my opinion, it is to present familiar things in a new light. In that, Karppinen succeeds remarkably well.
For a business example, let’s take a fresh case where we created a ChatGPT application for experts to use our company’s internal database. We figured that if the same application had been built two years ago, the price tag would have been around 100,000 euros. Now, the bill was just 20,000. So, the cost of the same implementation has dropped to a fifth over two years. Since the application significantly enhances the efficiency of work for all the company’s experts, we’re talking about days rather than months regarding the return on investment.
Many industries are currently facing a talent shortage, so why not employ tools that allow current employees to work much more efficiently? When machines handle machine tasks and humans do human tasks, perhaps we won’t drive people to exhaustion by the end of their shift. I want to believe that good companies can see these opportunities.
What should business management understand about AI and algorithms today?
- AI is already here today, and it’s not going anywhere. Things are going to change anyway, so it’s first and foremost about change management. How do you create the conditions for people to succeed in the midst of these changes?
- It would be a good idea to do a scenario exercise in which they think about what a competitor company in the same field could do by utilizing artificial intelligence.
- The decision-maker should know what would be the sweet spot of using artificial intelligence in his or hers own company. In other words, you should understand who in the company does repetitive work (for example text production and creating reports) and how much of that is done in general.
How should you get started?
Generally speaking, I’d say through something concrete. No Midjourney-created Avatars to represent the company on Instagram. It’s best to start from the everyday life of people.
Here are three concrete steps to start with.
- Company management must define how they use artificial intelligence tools at once. One example of this is instructing employees not to enter any customer data subject to the GDPR into ChatGPT. These require clear rules of the game.
- The company’s responsibility is also to offer its employees growth paths, i.e. to tell what artificial intelligence tools are and how they work, and in the future to also offer appropriate training for utilizing artificial intelligence. So learning should not be solely the responsibility of employees.
- As an individual, you should approach artificial intelligence with an open mind. Whenever you do some repetitive work, you might wonder if an artificial intelligence tool would be helpful.
Should there be more talk about AI at workplaces?
Based on my experiences, the topic – especially ChatGPT – has been widely discussed around coffee tables. It has also been thoroughly dissected in the media.
Next, I want to see us move from talk to trials. That more and more people are trying out AI tools in all sorts of tasks, such as in generating reports or crafting product descriptions.