Editor’s note:Harvard Business School ProfessorRembrand M. Koningstudies the impact of artificial intelligence on entrepreneurship. He co-authoreda recent paperexamining how tech startups view AI-related productivity gains, as well as a study comparing the structure of AI-native startups to other, comparable companies. We spoke with Koning about how AI is transforming startups, how it is impacting larger companies and how the technology could reshape traditional career paths across industries. The interview has been edited for length, clarity, and style.
Generally, what are you seeing from startups using AI?
Startups are a lot more nimble than big companies, so where they’re moving tells us something interesting about the future. Everybody is using AI tools, but the question is whether those tools are actually translating into productivity. If we look at AI-native startups, they need roughly 20% to 30% fewer workers than non-AI startups, which really surprised us. And these smaller firms are generating hundreds of thousands more in value per worker. Who they employ and how they manage those people also looks different.
You’re typically seeing flatter management structures?
Yes. In worrisome trends for MBAs everywhere, they seem to have fewer managers. And they seem to be hiring a greater share of technical talent. If, all of a sudden, we only need two people on a team who can use AI to get more and more types of work done, the need for a big, hierarchical organization structure with lots of managers coordinating talent diminishes. AI-Native firms can operate with the benefits of a flatter firm while producing more.
And you’ve also seen a shift toward more technical employees, even as AI can do technical work?
While AI helps with coding, it also can assist with marketing, sales, finance, and more. At least as of 2026, it appears there is still a need for technical knowledge to put AI into production. Particularly early on, if you’re one of these high-growth startups, you need your code to work. If your AI generated marketing campaign isn’t perfect, you’re going to survive. If your AI product doesn’t work, you’re toast. So, what we think is happening is that AI is allowing founders to delay hiring these other sorts of roles. They can use AI to come up with a sales playbook and delay hiring a salesperson longer than they could have without AI.
How does this translate to larger companies?
You’re likely not going to see layoffs, but larger firms might start to delay hiring or may simply not fill a role when somebody leaves. However, it’s really important to think about why we think these AI-native startups are smaller. Part of it is they’re using all the new tools—ChatGPT, Claude Code, Lovable—to be more efficient in how they work. But in our data it looks like the biggest gains come from putting AI in the product itself.
Take Gamma, a company that does AI-generated slide decks [and was the subject of a recent HBS Case co-authored by Koning]. In a sense, they’ve embedded a little “designer” into their product. As a result they have scaled to a $2 billion valuation, well over $100 million in annual recurring revenue and with roughly 50 employees by building around AI. Without AI, they likely would have had to hire tens of thousands of designers and ended up as a services businesses instead.
Returning to incumbents, I think the change will be more gradual, because it’s hard to completely re-engineer your company around AI like Gamma has done. Especially if you’re not a software company, it’s going to be even more complicated, because building around AI isn’t as straightforward. That said, I think incumbents should be thinking through both how they use AI in both ways, to supercharge their processes and to fundamentally rethink their products and services.
In your paper, you asked startup founders how much they’d have to increase headcount to replace AI. You found a mean of 56%, but the most common response was zero. What do you think accounts for this disparity?
We asked them the following hypothetical: Imagine if I took away all your AI tools, how many people would you need to hire to do the same amount of work at the same quality. What’s fun about this question is it captures the economic idea of productivity. If AI is enabling us to do more with the talent we have, this implies we would need more people to get our work done without this technology. We asked this question to high-growth startup founders and were a bit shocked to find that the most common answer (33% of respondents) was zero! Yet the second most common answer (at 10% of respondents) is that they’d have to at least double the size of their team to make up for the loss of AI. This was puzzling. Why did we see such big differences?
When we look at the data, it appears it has to do with how much the founder had formally integrated AI into their organization. Those that formally organized around AI got the biggest benefits, while those that just kind of gave their employees ChatGPT had much smaller benefits. I’m seeing this with a lot of startups in Silicon Valley. When new AI tools come out, they cancel their meetings and spend that day basically trying to reinvent their workflows, teaching everyone in the firm how to integrate this technology. When we look at big surveys of AI use, firms that use a lot of ChatGPT don’t report big productivity boosts because most firms don’t try to reinvent their workflows. However, when they do reimagine their work around AI, that is when we see the big productivity boosts.
For these leaner startups, as they get bigger, will they need to add salespeople and HR staff and all the familiar elements of bigger firms?
They’re definitely going to grow. Every time someone predicts the firm has finally disappeared, and met its match because of technology, they’ve been wrong. You’re going to get hierarchy, but I think the long term will be flatter. What’s interesting is we’re seeing a lot more demand for generalists in these firms, because AI can allow you to tackle lots of different sorts of problems. You’re also seeing everybody remaining an individual contributor, because even if you’re managing people, you’re also managing agents doing work on your behalf.
What are your thoughts on predictions this is going to have dire effects on employment?
These startups are 30% smaller, but the thing we can’t answer is how many more startups we will have. It’s unclear to me where we’re landing on that front. Maybe this allows smaller teams and more positions than in the past. When the App Store came along and made it easier to build mobile apps, it’s not like we got fewer mobile app developers. We actually got more, because developers built lots and lots of apps.
That said, I think it is important to acknowledge there is going to be creative destruction. It’s going to shift returns to different kinds of skills. If you’re a software engineer, your ability to know syntax really well and bang out lines of code is no longer going to set you apart. It will be different sorts of skills, and I think we’re going to see that in other professions as well.
Even at the consulting firms, you need to know vibe coding now. It’s going to be table stakes for every MBA around the world. Here at Harvard Business School, we’re training all our students now to use these tools as part of the first-year curriculum.
Another critical question is what it means for the traditional career ladder. A lot of the grunt work we used to do to gain judgment is starting to disappear. We are seeing tons of young entrepreneurs just out of college or their MBA launching new AI ventures. I think it’s harder to do within traditional organizations.
What might new career ladders look like?
I think they look a lot more entrepreneurial — a lot more people starting firms. I think we also need to potentially rethink what it means to do apprenticeship work and how we build judgment. If career ladders start to vanish, how can recent graduates acquire those skills without having to build things in the first place.
At BiGS, we welcome your input and invite you to engage with us. Please join us on LinkedIn and subscribe to our newsletter using the signup box at the bottom of this page. You may also reach us at bigsinst@hbs.edu to discuss collaboration or ideas on how business can play a role in addressing societal issues. We look forward to talking.
Actionable insights in your inbox
