You started your AI postgrad at Saïd in January 2026, so you’re about four months in now. Give us a picture of what an AI postgrad at Oxford in 2026 actually looks like from the inside. What surprised you, and what was different from what you expected?

It’s a postgrad program. The main goal is to give an idea of what AI is to people, to non-tech people like me. But we have a lot of data scientists and people from the IT industry in our cohort, which surprised me a lot, because the depth on the tech side is not as comprehensive as I would have thought those people would require. It’s covering different aspects of AI. It’s not only the tech part. It’s strategy, implementation, history, future outlook.

The most interesting part is the speakers they bring in. It’s just another level. People who are on the forefront of implementing AI in large organizations like Xerox, HP, the UK government, and so on. Some of the modules are more interesting for people with a tech background. The other ones are more interesting for us, for people who don’t have a tech background.

An interesting thing about this program, it is the most successful program at Oxford today. They are running two cohorts simultaneously, each 66 people. The maximum class size of another postgrad program is 40, and it’s only one cohort. We were selected from, they said, about 4,000 applicants. Around 120 people were selected to be in the two cohorts studying there.

What they’re trying to do is fill the class with people from different backgrounds. They want people from a tech background, from marketing, sales, startups, finance. You name it, we have all of them.

Another thing that surprised me, I expected my cohort to be quite comprehensive and to meet some brilliant people. But I’ll tell you, all 66 people, every single person is a very interesting person with a very interesting background. There are no people who are there by chance or just spending their time. Everyone there is steering their life with their own goals, where they feel AI is going to help them. Whether it’s their own business or their corporate role, a lot of them want to climb it faster, or they feel that this knowledge is going to help them be stronger.

They’re trying to prepare the leaders of the future, leaders who can implement AI in different domains, in business, in corporate life, in non-government organizations.

You went to Florida International in 2007 for business and legal English. That was pre-iPhone in any meaningful way, pre-ChatGPT by 15 years. You’re back in school for an AI postgrad nearly 20 years later. What is the difference between being a student then and being a student now? What can you do as a student today that the 2007 version of you could not have possibly done?

If I skim it down to one thing, the difference is that I can absorb much, much more information. Because in 2007, in Florida, if I had to prepare for an exam or a test or some kind of report, I had to go to the library and study for days, if not weeks, studying actual books.

Now, all the information in the world is right in front of me, in ChatGPT or Claude or Gemini. You can get a comprehensive summary on every topic. Even after Florida, when I went to Cranfield, which was another really challenging degree, Google Scholar was already there. But you still had to physically read through it and summarize yourself. Now it’s much easier to prepare yourself for any type of report.

For example, in Oxford recently, we were put in small groups of eight people, given a topic, and asked to come up with a business idea in seven minutes. In seven minutes, we prepared not only the presentation but the strategy, the marketing, the budget, the mission, the vision, everything. It’s mind-blowing how we change right now.

If I had to narrow it down to the one main thing, it’s the amount of information you can absorb and analyze in a given time.

The quality of your work is much higher. The level of your analysis and the level of the work you are delivering is much deeper, because you don’t have to go through the basics. Everything is at your fingertips. You just summarize or scan the inputs and then, using your human intelligence, you go deeper.

It’s much more interesting to study right now, because I’m not just doing the automatic stuff, the basics. I have to use my cognitive abilities much more to do the analysis and to come up with new ideas. It’s much more interesting.

It’s like you really have a smart assistant. You have an idea and you want to explore it deeper, and you start a very interesting conversation. How does it apply? Is it going to work if we implement it in this scenario? How is it going to play out if we, for example, have this much budget? Or if we have much less money? Or if we’re doing it in a different way? It’s so much more interesting. You feel like a real scientist when you do this job, because you are thinking through the process much more deeply.

AI and universities, there’s a war between them right now. Some have banned AI completely. There are issues of skill-development gaps, privacy, integrity. From the inside of Saïd at Oxford, tell us what AI use is actually like in the program. Is it required, is it banned, is it encouraged, is it ignored? How is it used day to day, inside the classroom and outside?

At this point, there isn’t any official policy from the University of Oxford on how you can use AI in your work or for your studies.

However, the professors are much more advanced than those policies. They’re like, guys, forget about those policies. I don’t want you to spend two days doing the referencing. I don’t want you to spend hours doing a table of contents. Leave it to AI.

I don’t want you to spend hours working on your language. Most of us are international students, and we don’t have the same language abilities as students from English-speaking countries. What they want to see is the depth. Because at this point it’s really obvious when someone is just generating their work and submitting it. There’s no depth in that kind of work. All the analysis is so shallow. If you’re paying all this money, it has to be meaningful. It’s not serious otherwise.

There are some limitations, of course. It’s a double-edged sword right now. They want you to produce work they’ll be proud of, work that means you’ll be out there in the future. At the same time, there’s so much going on right now with IP and other issues.

We have some professors who just tell us, guys, use AI as you want. I’m not going to frame you. I’m not going to give you any frames. Just go ahead, embrace it. I want to see how you think.

Some other professors are really cautious about it. They tell us, okay guys, you can use it for this, for this, for this. Try not to use it for this. Use it for the technical stuff. Use it for referencing, for structuring, for the table of contents. Use it for improving your writing. But don’t use it to write the substantial parts of your work.

Some other professors don’t mind if you use AI to write it, because the assignment itself pushes you to think. You just can’t give ChatGPT the topic and ask it to produce the work, because the work has to be really personal.

For example, when you’re writing about your personal development or the development of your own business. You can take your data, feed it to AI, run some analysis. But to analyze it effectively and to deliver really good, deep work, you need to use your cognitive abilities, your brain. Otherwise, it’s just going to be really poor work. It’s not going to work.

So yes, it depends on the professor. We have some guidelines from the university itself, but there is no software in the world right now that can fully detect AI-generated work. Enforcement is just impossible at this point.

I’ll tell you a short story about my friend who was studying at another UK university a couple of years ago, when ChatGPT was just released. He used AI to produce a substantial amount of his thesis. He got almost expelled from the university. He came to an agreement with the university that they would allow him to rewrite his thesis the next year and submit again.

When he went back to the university the next year to get the topic and meet his professor, they had completely changed the way you have to produce the work. The professor told him to use ChatGPT. Use it. Reference it. Show that you used it. And he was thinking, I just did it last year, and I got almost expelled because of it. Now you’re pressing me to use it. I didn’t know what was going on. It’s changing really fast.

There’s a very interesting thing going on in Oxford right now in how they’re transforming the education. Tell me about it.

Even institutions like Oxford feel that their bachelor degree programs are going to have a problem in a couple of years getting students. No one applies for these programs. They still haven’t figured out whether it’s because of technology or because the generation is different now. They are different. They don’t want to spend three or four years at the institution only studying. They want to do something else with their lives too. They want to work, travel. They have a hard time dedicating themselves to something a hundred percent.

So they are changing the way you receive your degrees in the future. It’s going to be a path you take, and you will have something like 10 years to complete it. You will get your degree as soon as you finish, for example, 10 programs within the university at your own pace. Whenever you want. If you want to do it in a classroom, come to Oxford. If you want to do it on the shore of the ocean, go ahead. It’s your choice.

You apply for a program, finish it, and it gives you credits. In a span of 10 years, when you’ve accumulated the credits they require for the bachelor’s degree, you get your degree. The same goes for the master’s degree. They don’t know what to do with the PhD yet, but PhD is still there because of the strong name of the university. There are still a lot of people who want to come and do a PhD there, because there are so many possibilities to get funded or get published when you’re doing your PhD at that kind of university.

So this is how education is transforming in the leading universities. I know Cambridge is doing the same right now, and it’s very interesting. I still have to do my research about the American universities to see how they are handling this problem.

We’ve gathered statements from faculty across this series. I’m going to give you five rapid-fire quotes from those conversations. I want your reactions, the thoughts that come when you hear them. As a student, do you agree? First: “Banning technology is like banning progress. Utilize AI, but this is where you have to lead, and not be led by AI.”

Totally agree. Totally agree. And I see it in my institution right now. You have to embrace it. You have to be first, or you’ll be left behind. Definitely.

“I am not all in on AI. I am all in on AI understanding. Oxford and the University of California system have secured access to frontier models for all their faculty, staff, and students because they see it as an equity issue.” Your reaction.

That’s a deep one. I agree that just using the tools is not enough. You have to understand how to use it. You have to understand the limitations of the machine. The machine is producing only the knowledge it has. It’s not creating anything new. And if you want to create anything new, to be on the forefront, you have to understand how these machines work.

“What we really need are smart young people who know how to work properly with AI and can have AI augment their skills. We are not really doing that very well, either in academia or in industry apprenticeship.”

I agree. We are calling it artificial intelligence, and the intelligence behind it is artificial. It has only the amount of knowledge we give it. So the way you are using and utilizing AI in your study or your work is completely your responsibility. You’re responsible for the outcome. You cannot throw the tasks over to the machine and expect it to do them for you. It’s your responsibility, and you have to understand how to use it to get good work.

“Some faculty asked, could we turn off AI on the campus? Fortunately, the answer is no. It’s educational malpractice not to train students in the effective use of AI.”

Totally agree. It’s one of the tools. Everybody’s going to use it. If you don’t have the skills to use it in the future, you’re going to be left behind.

“PhD students at Ivy League schools who used to know how to code and debug line by line cannot do it themselves anymore. And professors are not sure if that’s a problem or not.”

I don’t think that’s a problem, because it’s very mechanical work, it’s an automatic task. If you can delegate it to the machine and go deeper in your research, the quality of the PhD work is going to be much higher, in my opinion.

Our project is called Between Campus and Code, and you are literally between Campus and Code right now. Studying AI at Oxford while running Waterlok in Aparan. I assume that most decisions at Waterlok still happen on paper, in spreadsheets, or in someone’s head. Ten years out, in 2036, what does the right relationship between an Oxford-trained AI graduate and traditional Armenian industry actually look like? And what is the version of you ten years from now doing?

Very interesting question. I ask myself every single day.

When I came to this company, I was already really deep in AI. I was trying to implement it in different areas of my life, even my personal life. I was so into it that I was literally chatting with ChatGPT almost every single day before going to sleep. I was just amazed. I’d be going to sleep with a topic in my mind, for example, instead of watching a historical documentary, I’d just ask questions to ChatGPT and have a deep conversation. It’s just amazing.

And then I came to the company, and it was such a reality check for me. I came from Oxford, where we embrace every new technology and we try to automate everything, use the best possible tools. And I’m coming to a company where the director of production gives me a blank screen.

Everything was on paper, given to me every week. He had to go and manually calculate how much material he needs, how many labels, how many caps. I was just shocked. Even for an Armenian company, it was an outdated way of working.

And then we started. What I wanted to do was change everything at the same time, very fast, which never happens. So Oxford helped me design a strategy for how we were going to implement those changes in the organization, because it’s a big organization. We’re 120 people working here right now. It includes warehouse, production, distribution, logistics. It’s at Aparan, with 26 years of history.

We are now doing it step by step. The biggest pain point was, we were never ready for the season changing in Armenia. Last year, April was hotter than May. Because Armenia is a landlocked country, and we are blocked from both sides, so we cannot import materials fast.

So what we’re doing right now, we have a tool already, we are just testing it now. We’re implementing a tool that predicts our sales two weeks in advance, based on weather changes.

There’s significant information from the open weather data that we’re feeding to our model. We have 26 years of sales history. The weather data is there. It’s publicly available. We trained our model based on those patterns. And we’re feeding in the warehouse materials movement, feeding it the weather patterns going forward, and it gives us the outlook for how sales are going to look two weeks from now and what we need to buy to be prepared for it. And it works really well. Really well.

It’s just the beginning. There are so many things that can be done, especially in the manufacturing industry. So many things. Right now I’m concentrating on sales because I want to bring the revenue. I’ll be more relaxed next year, and I’ll go do it in production as well.

*This interview has been lightly edited for clarity and readability. The interviewee reviewed and approved the transcript before publication.