13 06 2026-PARTHENOPE- LUIGI CONGEDO DAL TASK AL PENSIERO CREATIVO

Okay, hi everyone. Good morning. I slightly changed the the title. So today we're going to talk about the cost of intelligence and what it means. But before we jump into the the this discussion, I'd like to introduce myself.

I'm All right, it's not working. All right, I'm Luigi Congedo. Uh I I'm from Napoli, but I live in US. I moved in San Francisco in 2012 and I was I'm a former venture capitalist that was lucky enough to get into the AI space 10 years ago when AI wasn't cool. I did 35 investment in AI with a company called Bootstrap Labs.

We have one IPO, seven exit, a lot of failures and I learned a lot about AI since day one, since AI was in the lab before it got to the mainstream market. In 2022, I decided to found a company called Data Masters inside of Italy. Our main office is in Bari. And the reality is that you know, we often talk about AI in terms of technology, investments, disruptions, but the number one important things that I think everyone should understand is that this is about talent, is about skills and is about collaboration between humans and machine. So what what we do is All right, I have some So what we do at Data Master is really we trying to build a platform to support professional and businesses with AI adoptions programs.

And we do that through three main services. Education, skills measurement, and development of AI solution. It's very important to understand that in the world that's moving so fast, you cannot invest in technology without investing in your people. Because often the number one problem for adoption it skills, it's not technology. We have served a little bit over you know I think at this point is 75,000 people over 100 companies so we work with a lot of Italian company multinationals.

The the we recently signed a contract with Paramount to help them to ship mobile apps for the European market with AI. Data Masters been you know recognized as one of the best at solution. I'm going to go fast on this one. I don't want to this to be a commercial. Let me jump into this so the AI impact.

First of all I would like to you to ask a question how many of you are using a laptop during your day or week of work? Everyone? Okay. And how many of you are using AI? So it's almost everyone but the reality is around the world is not the same.

We have about 3.5 billion people in the workforce. Only half of the people today are using a laptop, mobile phone or tablet for their jobs. So when it comes to AI adoption you need to understand that we are very privileged. We're we're being part of this revolution but not everyone else is.

And when it comes to adoption as I said skills is the number one issue. 45% of the people in the workforce among that 1.3 billion needs to reskill. Over 60% of the European company are claiming that they need trainings and they don't even find service providers that are good enough to train their people and most importantly what AI is doing is accelerating the skills obsolescence. On average what we learn today in two years and a half from now is going to be obsolescence.

And why this matters a lot to me is that you know we need to understand that AI is is not a single trend but it's the what I call the alpha trend. As As can see in this slide, uh you you have neural networks uh with a a slightly uh higher you know, bigger bubble. The reason is because every time AI gets better, every other industry gets better faster and change. AI is the interconnected in a way where we are advancing R&D, biology, cryptocurrency, automotive. So, this is not a just about ChatGPT getting better or Cloud releasing the next model.

This about acceleration. Even the COVID vaccine thing. COVID vaccine was invented in only few months because of AI. Because of our ability of processing data very fast. And when it comes to the cycle, often people ask me like, "Okay, where where are we in the cycle?"

AI did this cycle so many times already. Since 1950, humans are trying to mimic the our brain. So, we did this cycle so many times. And every time we do the cycle, what happened? We got excited.

Then few days later, we think AI doesn't work. And then eventually, some of the building blocks get adopted and creates productivity gains. But, you know, the most important things is to understand human intelligence and machine intelligence is not the same thing. We started by trying to reproduce the human brain, and we end up having a new intelligence. A new type of intelligence that can create uh that is start getting a lot of cognitive abilities like we used to have.

We were the only one we and animals were the only one with this cognitive abilities. Uh AI AI is also able to create now, to manage very complex processes, to become increasingly accessible. And accessibility is what happened in 2022 with ChatGPT. The world changed not because we had some disruptive inventions. The world changed because a company in California decided to create an interface with the most powerful machine in the world that was going through natural language processes.

Since then the adoption started, but the reality that that the core of this technology has been in the building for many many years. And most importantly, one of the core differences between our intelligence and machine intelligence, machine don't sleep. But I added this because it's a very important topic these days. Machine do run out of token because they're very expensive. And there are some risk for us as humans.

Laziness. Getting the wrong expectation. I ask the AI to do something and the AI doesn't do it. One example, we need to understand how to learn and learn very fast. And most importantly, we need to understand what the what is the level of autonomy versus independence we want to have in our life.

Some interesting trends that are happening today. First of all, the cognitive skills that I mentioned. Since 2017, machine are better than us at recognizing objects. So, this is a fact. They they can see.

And what it means, they can see the world. They can understand the world. It's not just about talking through through text. They understand the world. A camera here can understand who we are, what we do, our sex, our conversations, and everything thing everything else.

The second trend is that at the end of the day what's happening today is that humans are acting like machines and not the other way around. You know, this is an example of me or someone else asking ChatGPT, "What should I tell my wife to make her happy tonight?" That's the reality. A lot of people are doing that. They're asking AI what to do in the real world.

On the other side, which is even less intuitive, we are making decision like algorithms because we're getting so used. So, the next trip, the next restaurant, what you're going to do is like looking at the reviews, trying to collect data, and then you make the decision. So, we are already changing our behaviors. And you know, the multimodalities, as I said, vision, but AI is also getting better at producing images, speech, and so on. So, the modalities of inter- of interconnection and engagement and collaboration between us and machines is getting bigger and bigger and bigger and out of control in certain ways.

This is not about asking for an answer, but it's about understanding how to deal with a system that can plan, act, and collaborate with us. And I think you know what I will try to do today is go very fast on some of my cases that I've been developing in the last 10 years and and just to give you a All right. So, the number one thing is, you know, we already we all know this. A human versus an agent doing research and trying to understand the difference between act and the US rules on AI Act. The in agent 5 seconds, human operator 8 hours.

They're much better of us. But the reality of we're really going to ask AI to invent for us, to understand for us. And if if if this is the case, you know, at the end of the day, the the Number Okay. Okay. Now, yes.

That's better. So, so if it's the case the machine and if machine can create an invent, are we expecting AI to make all the new inventions instead of us? So, the real question is not if AI help us to make better job, but which part of the work remains human, which part is adjective, and specially who is capable of orchestrating this collaboration. Okay, guys. Sorry about the slide further.

I would love to go faster. Okay, here we go. First thesis I have, AI is not free. Agent Agents are changing the economics of automation. What it means.

I will ask, would you put high standing answering customer service for a restaurant? I don't think everyone will do that. It will It wouldn't make sense. But, this is what we're doing today. We're putting the most powerful models to do customer service.

And it's wrong. What is happening in the industry is that we went from AI everywhere in the last 5 years to AI by function until this year. You know, AI in sales, in finance, and blah blah But, then the reality the new approach is level three is right picking the right model the right for the right task and understand the right level of autonomy where humans needs to be involved. And this require a lot of people actually and a lot of expertise. So, the illusion of AI is going to do everything is wrong.

We started there, but we're now level three and this is going to get more and more complex. And I'm not talking about the the you know, like the infrastructure behind it. So, the cost of intelligence is getting higher. Every million of tokens cost more and the system the big tech are giving us they're do so many complex operation. They're cost us so much money.

The number one problem for businesses today is not firing, it's not hiring, it's running out of budget. Because they put AI everywhere and now they're scaling back. Number two, junior skills are more exposed and that's true. They're getting out of the market because they're not thinking. They have the wrong expectation.

You know, I remember you know, everyone we were kids falling down the stairs. You You hit your head, learn, don't do that again. Or if you're very smart, do what this kid is doing. Gear up, do that again. And you're not going to hurt yourself.

But, the problem is not to do the job. The problem is to learn how to do the job in fact. And you know, juniors are typically good at researching, analyzing, documentation, writing the first draft. No one does that anymore. But, that requires a lot of thinking, right?

So, AI is getting better at doing the task, and less people are getting hired. It's true. 50% of the task are complete by AI. It's true. But, where do the junior people fit in the industry?

They need to evolve from junior executor of single task to AI orchestrator. Understand okay, which step of the process has the AI uh you know, is the AI better than me? How to design a workflow? And most importantly, evaluation of the output. We cannot rely on the AI to create the output.

We can ask the AI to do it, but we need to review it. And this is where junior people can spend a lot of time and and do a very good job. And of course, as my um colleague Milly was saying, continuous learning. This is number three. Senior skills become more important.

You know, this is an example of a senior developer turning into chief problem officer. AI at the end of the day is about productivity. It's about solving problem. And what is very interesting is that historically senior people would not code that much, but today there's an increase of senior people shipping code. Why?

Because they're very good orchestrators. They understand the business. They can have an impact. And most importantly, they're going to have the trust and the budget for their companies. And you know, the premium skill, 56% increase in salaries for people that know how to do the job.

So, when you look at the media, everyone talks about company firing. Company are not firing because of AI, guys. Company are firing because they did the wrong math. During COVID, they expect the world to go digital at a very fast pace. They hired too much, and they're scaling back.

Because if you look at the actual data since, you know, beginning of last year, because of my slide, AI everywhere, AI been factored, and and level three, where we are now, company need more people. They are realizing that because the cost of intelligence is going up and we need to rely on our intelligence. This is number four, creativity doesn't disappear, it moves up to the value chain. This is very important. As humans, we have a very big responsibility.

Because we become orchestrator, we become problem solver, and we move up to the value chain. We're not doing the first draft anymore, we're not doing analysis anymore, but we do need to make that decision. We need We do need to differentiate our decisions for our competitors. And this this already happened in the financial industries. You know, financial industries adopted algorithms way before any other industries.

But eventually, what they realize is that, "Oh wow, now my algorithms is giving me the same of my competitor algorithms." Look about how many people got hired in the financial services in the last 15 years. The industry was is is is crazy. It It become like probably like five, six order of magnitude bigger. Because as the competition races, you need more creativity to differentiate from your competitors.

And you know, we should be optimistic. I think yes, I'm a paranoid optimistic because to be an entrepreneur, you have to be. But are we going to be be replaced? Well, we we weren't replaced in the agricultural revolution. And if Well, when people ask me, "Are we going to be dependent?"

But let's be honest. on innovation and technology. So, I'm not too worried about that. I'm not too worried about replacing. I'm just really focused on on, you know, where we fit into this equation.

And you know, it's normal that at the beginning of this revolution, which is as I said, is an alpha trend impacting everything, we're scared. And just out of curiosity, how many of you know where the word sabotage came from? Does anyone knows? The shoes. And why would you sabotage for us as sabotages today?

Do you know it? Okay, so sabotage is a shoes, wood made shoes that people use especially in North Europe. The word sabotage as we use as we use it today comes from the industrial revolution because the worker were going and night into the fabrics into the factories taking their shoes and throwing them at the machines to stop them because they were afraid. And this is what we're doing right now in Europe. We're trying to stop progress.

And we're not realizing that this is not a war. AI could be our friends if we participate in the making. So AI generates possibility, humans needs to define direction, meaning, intentions. And when I talk about elevating human creativity, I want to be very clear. This is not some slogan.

It's not another Oh, yeah, let's be optimistic. There's some very specific items of which we can have an impact as humans. Strategic direction, intent, prioritization, judgement, ethical values, meaning and context. For instance, most of the work for those who building AI these days, it was prompting at the beginning and now it's called context. We need to give context to the machine in order to make the machine work and add value.

And at the end of the day, the ultimate responsibility of the impact, it is on us, not on the machine. And I'm going to you know, try to go to the next slide. Uh here we go. Uh and so, you know, at the end of the day what I'd like to say that this is not a war. It's a it's a new architecture.

It's a new social order maybe. I like to call it architecture. You know, AI and the partnership between humans and machines is redefining the human role in the workspace as well in the society. We are moving up the value and need to do higher thinking. Have you know, everyone is to own their decision because once you become an orchestrator your decision has a a way bigger impact that doing an analysis, writing the first draft, and so on.

So, if you want to participate in the market, you need to make sure you have ownership. You have curiosity. You love what you do. You exactly know the problem and talk to other people. Because otherwise you're getting the wrong expectation.

And just to close with something I love you know, when we talk about innovation cycle and AI and why and how we get there. You know, just jumping back into 1960 where the semiconductor industry started, the world start changing. And we, you know, went from 1960 to 1990 system integrator, the internet, the dot com. Back then, what happened? Back then most of the world was predominantly not using technology and most of the people making technology were tech people with a lot of years of studies and and so the participation was very low.

19 to 2020, what happened with internet mobile, everyone was connected thanks to the mobile product. There are still many people that connected the first time thanks to a smartphone and not a laptop. And what happened between 20 and 22 with the cloud also is that software was used by almost everyone, even people without tech skills. And that was the paradigm switch. Everyone start using it.

But then in 2022 where GPT got out, a new change in behavior and participation happened, which is as I said, until few years ago everyone was using software. Now everyone can make software. Even people without deep technical knowledge can participate in the making of the software thanks to AI. And I think we need to decide whether we want to see this as a painkiller to solve some problem or a vitamin to add our value, our wealth, our own prosperity and I hope our society. Thank you everyone.