CTRL/SHIFT - 13 06 2026 CHAPEL STAGE - GIANLUCA DI BELLA - CLAUDE K*ED MY JOB I'VE NEVER BEEN BUSIER

Okay, everyone. This speech is actually a bit of a joke trying to make and touch actually a serious point. You know, nowadays AI is everywhere and also it's quite strange that everybody thinks that AI is taking our job. So, I'm doing this talk to let you understand that actually AI is just changing the way how we work, but humans are still needed and AI will never take the creativity and all the visions that a human can have. So, as you can see the title is a bit provocative and it's a cloud killed my job and I've never been busier because actually if we made the work for free, why we can't we can't stop working?

Let's start with a little intro about myself. Let's see if this works. Okay, so without trying to be so much egocentric, I am Gianluca Di Bella and founder of this event, one of the co-founder and I'm also founder and CEO of Mood Studio. We are a nerds dev studio that basically can build anything and you know, we are full of designers, full of developers. We work with the Italian Space Agency, banks, you know, different type of stuff and as you can understand as a as a developer, AI is in in my workflow everyday.

So, you know, AI is actually out from a few years, but you have to understand that I never work like that. Like you know, my day is just busy, busy more everyday just because AI give us the permission to work more. So, that's the real topic that I want to touch, right? That the AI just give the opportunity to work more because normal stuff can be automated without caring about boring and continuous task. We can focus on management, we can focus on different type of stuff that before could could take us a lot of time.

So, you know, this is basically the last two years of our life. Every day I wake up, I open Instagram, I open X. Uh internet has already buried us, literally. Our devs dead. Let's prepare for the AI job apocalypse.

Let's the vibe trap. Software developers are are not working anymore. So, basically, you know, it's kind of like, you know, really complicated for people because imagine like a guy that is at university at 18 years old and is studying. He's like, why should I be motivated to to work or to study if the AI can do everything for me? But the real stuff is actually that we need to change mindset because we still need to study, we still need to understand how actually life works and all the economy around us trying to use this AI as a tool.

So, let's let's go a little bit to understand the data of this AI implementation. Let's see if actually it's changing in terms of of the labor market that we have around. So, let's see in terms of data what actually is happening. So, I think that everybody use at least once a day ChatGPT, Claude, or whatever. But if we see the data and I took those data from Statista, from different platform that we can find online, and you have to think that just 23% of the of our of the people in Europe, so these are data based in Europe, just 23% use AI.

You have to think that Instagram is used from 70% of the European population. So, there is like a crazy gap still. So, we think that everybody use AI, but actually this is not true. Um while actually from from the side of the B2B, from the corporate side, the growth and adoption of AI is way faster because already the 86% of companies expect that by 2030 all their employees will use AI. Of course, an entry-level position is really exposed to that because because um Let's imagine that, you know, I'm a CEO of a company and I have to hire a new developer.

Now, I can hire a senior developer without actually spending so much money in terms of salary because they can work less thanks to AI. But for entry position, for entry-level position, this is quite different because interns are the real problem of this supply chain of the workflow because nowadays entry-level positions are more exposed because what an intern can do can be easily done from an AI. So, it's a bit like where actually the supply chain of the workflow is touching. What is the main issue in the in in the workflow, yeah? So, actually like okay, AI is amazing, right?

So, what is my point here? AI is amazing. It's really productive. We can express so many stuff using it. But the problem that almost nobody is talking about the bill.

AI is really expensive. Like of course, if you use ChatGPT to make a document or a presentation, it's quite cheap because you prompt one time and you have done it. But if you're actually a big company and want to implement AI for all your workflow and for all the stuff that normally, let's say, compliance check, audits, fully development chain, it's really complicated. For example, in my company last month we paid 3,000 3,000 euro of tokens. But I can hire a developer for the same amount of money.

So, the thing is that okay, we can use AI, but we need always human support because human plus AI is a really shift in the labor market. Just AI can never substitute the human chain. Uh and actually like this is quite quite funny because in just like two years so because there is a big difference between ChatGPT and AI Agentech, right? Because Agentech coding is like every time you use cloud code, every prompt you're spending like let's say even 20 cents. And imagine like these this ratio of computational for for large corporations.

So basically like just 26 just 26% of companies fully understand their AI spending. So they are in a discovery phase. They try to understand if AI can can actually help them or not. So this is really a test that we are living right now. And of course like you can see a little chart of how the spending of VIs is growing and this will be more and more and more over year, of course.

Let's see the Uber case. Everybody know Uber because it's I mean a renting company. I mean I live in Amsterdam so you can imagine how many times we use it. AI enables a bit different. But you know, Uber Uber actually assign a budget of around 4 billion for 2026 for AI spending.

In the just the just the first six months, now we are in June, they already spent almost all of the budget. And they said, "Okay, I'm not going to use AI anymore because it's so much expensive." So like the real point is is actually to understand where we can use AI because I think there was a bit of AI washing. I would like to call it like that. That all the companies and all the people say, "Okay, I can do everything with the AI so let's use it."

But in the end of the day there are some tasks that you don't require AI that is so expensive that it's better to use normal labor ways of working, right? And that's there are so many cases from from Uber. There is a nice article on Reuters that you can check about it. And it's it's quite funny because the micro Microsoft CEO, you have to think that you know, all these companies, you know, American markets are quite funny because sorry, yeah. I'm sorry for that.

I'm also organizer of this event, so it's it's a bit quite busy. Um so, uh yeah, basically like I was saying that, you know, back in the days Microsoft invested in OpenAI, right? And and then they invested also in in Claude, even if the first investor in Anthropic was Amazon. And they said, "Okay, let's try to use both." So, they they use OpenAI and Codex for so many computational task, while they use more Claude code for management related task.

And basically, last month the biggest company in the world, Microsoft, that should support the AI, canceled all the Claude code subscription for all the employees. They said, "Let's not use Claude code anymore. Let's come back to code in a normal way. Let's use it just in the inception of the coding supply chain, not for everything." So, what does that mean?

You know, if a company of a size of Microsoft start to cancel Claude code internally, is that, you know, is not that the AI is not good, that the tool didn't fail. It actually got got fired because it's too good, you know? We cannot like accelerate too much the supply chain work. We need to be cautious into that. And that's why we saw now that US government this morning stopped to use Mitosis 5, stopped to use Fable 5.

And also like Anthropic last week, supported by the Pope, and in this stage at 5:00 p.m. we're doing a panel about that, about AI ethics. It's basically that they said, "Guys, let's stop because if you work in this sector, you know it. Every week there is a new model.

Every week there is a new news. It's so It's so fast that it's impossible to track even for the people that are working in this industry. It's just impossible. Like me, I'm a developer and every week there is a new model. There is a a new stuff to try.

And even for them, it's complicated to take track. So, okay, let's embrace it. But if there is no point, we don't understand how we embrace it. Let's not continue the research. For example, the main problem about that is the generative AI flow.

Because nowadays cloud code engineers are using are using AI to regenerate the code itself. So, what does that mean? AI is being coded from itself. Like cloud code which we are still going They were coding all of them. Now, the majority of the software and the and the and the platform from cloud code is built from AI.

So, we are entering this this phenomenon. If you're interested, it's called the Homer loop as Homer Simpson. And it's basically like this Homer loop is something that AI is continuing generated itself. And if we don't stop that, it's quite dangerous, right? Because we saw with Mitosis 5, the president of European Commission Lagarde called all the European banks, Santander, Cell, Intesa and all the you know, UBS, all the top European banks and said, "Guys, let's stop it."

You know, because if Mitosis 5 is out, our security is not ready. The level of security that we have thanks to the law and thanks to all other stuff is not ready for this AI adoption. So, I'm really scary because if we arrive to the point that AI is generating itself, there is no point to stop it. So, that's why and remember this word, mark my words, in the next months, we are seeing a stop of it. We are seeing a stop of the evolution of the eye.

So, we say, "Okay, we have a Hopus 4.8. We have a Mitus 5. Let's use it. Let's understand how we can implement it without going into accelerate every model every week."

Because it was a race, right? It was like, "Okay, let's create the best model. Let's compete because that's capitalism." We saw every week ChatGPT 5.5, Claude 4.

8, Gemini 3.3. Every week there is a new model. But, it's like it's more like researching on a new model instead of researching how to implement in our actual life this model. So, I'm I'm really scared about one thing about that that I just thought about is that, you know, if US is stopping because Western society care about diplomacy, democracy, and blah blah blah.

If China doesn't stop, that's a problem. Because if you see this race and to accelerate the AI development is because US needs to and Western society in general needs to dominate against China. And they have way better electronics electric grid. They have way better computational scale than us. So, if we stop to actually understand how to implement the AI, might be dangerous because China and general Chinese company can actually have a better model than us.

So, it's really complicated, you know, what we should do. We should actually stop to try to try to understand how to implement the AI or we should continue to accelerate to accelerate development and the research to always compete against China. And this is called the the Jevons paradox that actually was said from Nadella, the CEO of Microsoft back in the days in 2025. You know, it's one you cannot control anymore the the machine. So, yeah, I mean, I I'm like a radical optimistic in that sense.

So, I I'm really I'm really positive about, you know, the development and our future. But, one thing that I'm scared is like is is this competition competition actually healthy or not? So, yeah, I'm really curious about the rule of the of the government here about the the rule of the church, how they actually can stop and understand how the AI ethics can work around us. And um So, you can see like a little bit of other data about how actually like AI is being implemented. And you can see that like basically there is no real point because on the left we saw a chart that distributions of implementation where was was really fast, right?

In the back recently. But, also like the real implement like the the training of employees was lower. So, we had like a crazy implementation, but the real training in the employees in the in the work chain was lower. So, there is not a real implementation. Like, for example, if I'm a software house, of course I'm I'm using AI.

But, if you talk with big companies, they are still not using. For example, like all the all the American companies that have Microsoft Teams that I personally ate, they use Copilot. By law, because you know, for private data you cannot put like, you know, company data on cloud or ChatGPT. So, you have to use Copilot and Copilot is is not so great, right? And there is also obvious loop.

So, what does that mean that innovation goes crazy every model every new week, but real implementation is not happening. That's quite a paradox, right? And then like, you know, in my opinion like everybody's saying, "Okay, AI is stealing my job. You know, we are dead." and so on.

But, actually like I think that just the labor market shifted, you know? Nothing is actually died, just something moved, you know? So, basically like, you know, before we were like typing, writing emails, writing essays, and so on. And so, I think that like the task that were really human based as typing, pixel pushing, drafty drafting are completely dead, right? But now there is like a lot about the human taste.

So judgement, taste, accountability, that's something that the eye cannot do, right? And so basically like I like to say this also to all my partners and friends that actually clients stopped paying us to write code, but they pay us to decide which code to write. So the human rule is more like as a advisor as a strategy a strategy than just like a slave that needs to code there for people. So like you know, I really see the eye implementation as like a new industrial revolution. That's actually like you know, I love blockchain.

I love like you know, innovation in general, but I think that the eye is a different stuff. It's like you know, it's literally we can compare it to the to the first fire or to the steam or like to the electricity. So I'm I feel lucky. Like every day I wake up and I'm like my God, what a beautiful life we can live because we can do everything. Like there is no limits.

Like literally with the eye we don't have limits anymore. So you know, like I can see I can compare this this scale from we had we went from the steam to the electricity to the computer and now we have the eyes. So let's say that as the you know, the usual example of like from papers to emails and so on. In terms of like let's say finance, ATMs didn't kill bank tellers, right? They just changed it.

Spreadsheets didn't kill accountants. They just moved them up. So I think that just we're living a new evolution. But not just like in terms of technology. As a human race, we are evolving, you know?

We are becoming smarter and smarter thanks to the eye. And you know, like actually three reasons why we can stop as a society. Demand exploded and even even like as a business as a software house, I can say like we never worked like that. Because if a company had an issue to implement something, now they can write the AI and they can have a clear way to do that. If you are a normal dev with the AI, you're a senior dev, right?

So demand demand exploded. Every business that couldn't do a for software now can. Ambition inflated. For example, like we used to ship one product from for for one month, now we can actually ship so many products for just for one month, right? So the work is a bit like it's quite invisible.

So you know, it's a bit like as again, mainly advisory. We don't write the code, but we decide and we review it. So this is my favorite slide. It's basically like actually understand how new jobs are being born, right? Like it's basically a new shift and I think that now we should build things for AI.

Like before we were building a website for humans and we were building like marketplace for AI. We need to understand that we are not alone anymore. There is another type of I don't know creator how can I say that that actually is living the web world with us, you know? If you see Twitter users, 40% are AI agents. Nowadays meta just got fined from the European Commission because 30% of users were bots.

So of course, without control we are completely up because it's you know, I mean, what is the point to have a social network for agents? Actually it's also out there. But the point is just that way that we need to coexist and we need to co-live with agents. And that's the new evolution that is happening around us. So you know, this is so funny because without the AI you you can be basically like you know, you we have a junior position, middle position and senior.

With With the AI, if you're a new a junior a junior, you can become a normal dev. If you're a middle dev, you can become a senior dev. If you're a senior, you can become like a CEO or something, right? So, the level of intelligence of of general people just got crazy high, you know? It just acts up.

And that's the power that we need to use. It's a tool that permit us to do everything. That's it. There are no limits anymore. And uh you know, like uh let's be a little bit, you know, positive also because everybody say, "Ah, there are no job position anymore.

Uh there are cuts and so on." And that was true because when ChatGPT was out in 2022, 2023, all the companies were stopping to hire. What's happening now? They start to hire again because they realize that they cannot do everything with the AI agents. So, for young generations, like let's continue to go to university.

Let them continue to study because like we still need humans, you know? It's not true because we can see it. Like, you know, there are literally World Economic Forum data of last month. And also on on last meeting on in Davos, you know, like it's just that the future of jobs report, there are 78 million new job opportunities. They are creating new jobs, you know?

That's That's just a shift in the labor market. Nothing died. It's just a shift. So, like, you know, just to understand how in my opinion the mentality should work here. Like, we need to stop to competing with the machine.

So, we don't we don't have to be jealous against them that they can do uh better things than us. But, we need to start direct it, you know? And we need to start from the law. We need to start from the school, from the high schools, you know? Instead of having like we are in a church, you know, with all due respect in the religion.

It's Instead of having the religion hour in your school, you should have like also the the AI the AI uh Uh, AI hour, you know? You should like do coding class. People should code because before the the nerds were seeing like us disguised that were like in the black room coding and whatever, but now with the AI everybody can code. So, there is like this shift of paradigm that the coding should be a general subject that everybody should understand. So, you know, um, I'm quite optimistic, of course, but my message that I want to I want pass to you is actually that, you know, there is not unemployment, there is a promotion for the human, you know?

We should like be glad of what is happening because there are no limits anymore. So, you know, I don't I don't want to continue so much because, of course, I want to leave the the space to other speakers and continue with our agenda. And, you know, if you are curious to make any questions, you can come to me. Uh, I think that almost of you know me, so it's quite easy to discuss about those topics, but, you know, before that you go before that you go actually like sit with this question, you know, if the machine does the work, what are actually paid for for for the work that we are doing now, you know? If your AI bill is becoming a salary line, who owns the budget, right?

What is happening in terms of cost? So, have a look on those questions and actually repeat them yourself every day every day. How can I use AI, you know? But still as a tool, not as a a substitute. So, thank you very much.

Uh, that's my QR code if you want to talk with me. And enjoy the event, of course. I hope that was not so boring. I tried to be a little bit, uh, satiric in that sense. So, if you want to discuss and, you know, how we can implement the AI in general, how we can learn the on top AI implementation, you can scan the QR code and, uh, we can meet later in the aperitif in the agenda.

Thank you so much.