Generative AI and Business World in Turkey

An overview of the generative AI ecosystem in Turkey, startup investments, and corporate adoption trends.

Speakers: Zeynep Mavi

Content author

22 Nov 2023

Video Transcript

⏱ 14 min read• 2,397 words

Read the full speech or switch languages

I think you can hear me. Hi everyone, I'm Zeynep Mavi. I'm with the Turkish Artificial Intelligence Initiative. Today I'll be talking about Generative AI and business in Turkey. First, let me quickly cover what we do at the Turkish Artificial Intelligence Initiative. We were founded in 2017. Here's why we exist. As you know, AI is moving very fast worldwide. We didn't want Turkey falling behind. So we set out to build AI awareness and grow the ecosystem here. Over the past 7 years, we built that ecosystem. When we started, it was really tiny. Today we have over 300 AI startups that passed our validation process. Back at the start, that number was just 5 or 6.

All of our current startup members have gone through validation. By the way. If you have an AI startup and you're not in the ecosystem. Reach out to us. We update our startup map every 3 months. That covers new startups, pivots, and shutdowns. Our map is also used as an inventory by the Presidency's Digital Transformation Office. We have academic partners too. We work with more than 20 universities and over 300 academics. Most of these academics specialize in AI. We have local and international tech partners. There's the Tray Community. It includes NGOs, professional student clubs, and AI centers too. We also have our scale-ups. Those are basically startups that grew up. And we have our supporters.

Every company you see here has started some kind of AI journey. Of course, they're not all at the same stage. Some are really far along. They build their own AI tech. They have their own AI teams. But others just started. They say, "We have data. We want to make sense of it and use it." Those companies are at the very beginning. We leaned heavily on these companies for the Generative AI report I'll mention shortly. These companies make up a big chunk of Turkey's AI ecosystem right now. And our stakeholder list grows every day. Companies in Turkey keep realizing AI is essential. So honestly, new companies call us every single day.

Everyone asks us the same thing. "We want to start our AI journey. But where do we begin? We don't know what to do." So we draw them a roadmap to guide the way. That's why they reach out. And most importantly, we bring them into the ecosystem. At the Turkish Artificial Intelligence Initiative. We have a generative AI report coming out in October. That's what I'll be talking about today. It has some findings. Things we've seen about Türkiye. The report has a few parts.

First, a history of generative AI. Then 20 key terms. Then the transformative impact of generative AI on business. That part covers use cases companies can run, with examples from around the world. Today I'll talk about generative AI in Türkiye. How the private sector here approaches generative AI tech. And the generative AI startups in Türkiye. As I mentioned, we ran this survey across the companies you see here. The Presidency's Digital Transformation Office backed it too. So it also went out to companies they approved. This is really just a starting point.

Company size is generally between 1,000 and 5,000 employees. We have examples from a lot of different sectors. We try to show AI adoption in every sector we can. Here we have automotive, finance, energy, manufacturing, insurance, telecom, and FMCG. Our first question was simple. "Do you use generative AI at your company?" 64.5% said "Yes. We use it." 33.5% said "No. But we plan to in the future." And only 2% said "No. And we're undecided about using it anytime soon.". First off, this data made us really happy. Because Generative AI is a genuinely big deal. We hear this a lot, for example: "Is Generative AI a bubble? When does it pop?" Like, we heard about the Metaverse and then never again. What happens here? Generative AI will be gone next year, that kind of thing.

That's not really a thing. Generative AI is here to stay. It has a serious transformative impact. And that keeps growing. We'll see many more examples of it going forward. As you can see, most companies in our ecosystem know this. That's why they've already put it to work. So how are they using it? Generative AI has a ton of use cases, by the way. Here we showed the three companies pick most: content creation, R&D work, and process automation. Content creation and editing is, as we all know. The most common use of Generative AI right now.

On the R&D side, we see a big shift at companies right now. Companies used to build their own tech with only their own teams. Now they're moving to open innovation. They partner with third parties to adopt new tech much faster. Take Generative AI startups, for example. Partnerships built around Generative AI tools really speed up their processes. So what's the role of generative AI tech inside these companies? Honestly, I think this is one of the most important data points. Because here we're not just seeing Generative AI. We're seeing exactly where AI should sit. Let me open a quick parenthesis. We run AI awareness seminars. In those seminars, we always start with this. AI isn't here to fire anyone. It's not here to shut down any department.

Think about your daily grunt work. Some of it can be automated. AI is here to clear all that out. It's here to make you more productive. Nobody has to lose their job because of AI. And it shouldn't work that way either. And we see it in these companies too. Only 3.2% said it's at the center of our strategy. 16.1% said it's a big part, but not the center. Most important: more than half called it a tool that supports their strategy. That's really what AI is.

Your daily work, your data. It's here to support all of it. So seeing that data made us happy. So why do people want to invest in Generative AI right now? First, automating business processes. No question. Like I said a minute ago, your employees' time is valuable. Your own time is too. At some point, that's a financial resource. So automating processes matters. The time you save, the efficiency you gain. Those are big reasons to invest. Cutting costs is really tangled up with that same thing. You might need fewer people on staff. Or let me give you the simplest example. Say it's a Generative AI case. You need video, music, and visuals for a piece of content.

Instead of the fees you'd pay an agency, you produce it with Generative AI. Your cost drops way lower. Boosting your data and analytics capacity is another example. Sure, these are great reasons to invest. So then everybody should just invest in Generative AI, right? It's not that simple, of course. These investment decisions have a few factors. First, return on investment. As much as we support every company integrating AI. It's not possible for all of them. Because every company's operations, structure, and team makeup are very different.

So the return on investment varies from company to company. You have to take that into account here. How much sense does this Generative AI investment actually make for you? Same goes for integrating it into your business processes. Automation can be a huge time-saver for your workflows. But it can be really hard too. And then there's the tech infrastructure you'd need. All of this has to be reviewed very seriously in investment decisions. Of course, there are challenges. Fine, we decided on Generative AI. We'll make the investment. But along the way, we run into integration problems. First, security and privacy concerns. I want to pass this note along to everyone. Because we've heard this from a surprising number of companies.

We were at an AI awareness seminar again, three to five months ago. We were at a genuinely well-known Turkish bank. And we started chatting about ChatGPT and so on. They told us: "Oh, we use ChatGPT every day. We use it for everything." We were like: "You can't be using it for everything. You're a bank, after all. Your data is really valuable." They said: "No, we use it for everything. We get our work done with it." And we said: "Please don't. Don't share your company information. It isn't secure."

There's just no such thing in ChatGPT. Don't share your company information. Nothing guarantees it stays confidential. And beyond that, it has its own security holes. Sure, using it to produce simple work is great. Generating routine stuff is fine. But security and privacy concerns absolutely need to come first. They have to be addressed. Lack of adequate resources and budget is another big one. Because when you want to integrate Generative AI, you'll need a technical team for it. They'll get it and plug it into their own systems. There may be gaps there. Same goes for investment. Lack of access to good data sets came up too. Our survey respondents flagged it. So what are these companies doing about integration?

First off, most want to build the capability in house. They don't want to hand it to third parties. They'd rather train their own teams and build their own tech. 34% see partnering with third parties as the healthier path right now. Another 16.2% haven't integrated anything yet. They're just using off-the-shelf Generative AI tools. Before I move on, this data gave us one more takeaway. I should add it here. We talked separately with the companies building in house. What do you want to do, and how? If I remember right, 52% run internal trainings today. That's company training on Generative AI. The rest bring those trainings in from outside. They're buying consulting too.

And they really need it right now. No question. Training topics include intro to Generative AI. Also Generative AI prompt writing, safe Generative AI use, and more. So what about generative AI startups in Turkey? What are they up to? First off, we've got 6 generative AI startups in our ecosystem right now. They're Creative, ConsciousLab, MIT AI, Apollon Engine, Novus Writer, and Yazmatik. Since they're Generative AI, you could call them the apple of our eye. We started out with just 5 AI startups. Now we're past 300. And right now we have 6 generative AI startups. We really hope that number grows soon. We're backing them as much as we can.

Each of these startups tackles a different area. Apollon Engine, for example, handles music generation. MIT AI does pattern generation with Generative AI. That's why we introduced them to some great retail brands. Novus Writer helps companies with content production. Yazmatik does the same on the sentence and copy generation side. I'd also like to talk a bit about our other companies. We ran some interviews with companies that have integrated Generative AI. Let me share what came out of those. First up, Turkcell. On Turkcell's side, they take the data they collect from customers. They're planning to build a customer dashboard from it. It'll be organized by topic and served right to the customer.

They also have plans for a chatbot built with Generative AI. It would help new hires through onboarding. Honestly, this line from their interview really stuck with me. We're aiming to move faster on adapting our existing solutions. We want to step into the AI era one step ahead. Today we only got a preview of it. By the way, I can say the same thing about every company I'll mention here. They're all moving really fast right now. They're testing Generative AI through trial and error. How can they plug it into their systems? What results can they get? That's what they're trying to find out.

Now for Arçelik. On Generative AI, Arçelik is easily the most talked-about company in the ecosystem. We actually ran a Generative AI panel about a month ago. We had speakers from the Presidency's Digital Transformation Office, Microsoft, Arçelik, and TÜBİTAK. They walked through their own systems in more detail there. Arçelik has more than 100 Generative AI use cases right now. They also have teammates inside the teams called AI Champions. And with those AI Champions, they're building a ready-to-use Generative AI prompt library. They keep testing their prompts quickly. But even Arçelik, with the most use cases right now, says the same thing. There's still very little know-how in the ecosystem. You have to discover everything yourself. Right now that's exactly what every company is racing to do.

Over at Yapı Kredi Teknoloji, they're a bank. So their side is obviously harder. Especially around data. Data security, specifically. KVKK means customer data can't be shared. That narrowed our use cases. We also run separate work to remove personal data customers share themselves. Right now Yapı Kredi Teknoloji's biggest challenge is data security and privacy. They're steering their work around that. Those were the four companies we picked for interviews. And all four interviews had one thing in common. We asked every one of them: "What value have you gotten with Generative AI? How's it going? What results do you have so far?" The answer was the same across the board: "We're very hopeful. We're doing great work. But it's still too early to talk about results."

We'll keep tracking their progress. And we'll share it with the ecosystem too. Let me wrap up with a general summary. This is my last slide, I think. I've walked you through the report. So what does it tell us? What did we get out of it? In Turkey, every company is really starting to grasp how essential AI is. We're very happy about that. And the added value Generative AI brings is landing too. We'll see that much more clearly in the coming years.

If any of you run your own company, I definitely recommend this. Start using Generative AI in your company. And of course. If you have an AI startup or a Generative AI startup. I'd love to see you in the ecosystem. It could be our Turkey AI Initiative, or the Turkey AI Ecosystem. If you have any questions about these, or want more info, reach out to me. Thank you.

This transcript was generated using AI speech recognition and verified against subtitle style rules.

Content author

Zeo's Guest

author

Contact us

How we handle your data

Open as a full page