Artificial intelligence: What awaits Turkey in 2024? – DW – 24/12/2023

One of the most discussed topics of 2023 in the world was productive artificial intelligence technologies. Owned by OpenAI ChatGPT Since its launch in November 2022, interest and investment in productive artificial intelligence technologies has increased. Productive AI technologies that underpin products like ChatGPT, Midjourney, DALL-E, Google Bard have now become an assistant that writes, draws, produces video content on various topics, and speaks like a human.

The point where the giant tech companies now want to get to is to create a human-level general intelligence called “AGI” (Artificial General Intelligence). Every day there are new reports of artificial intelligence technologies that have the capacity to solve mathematical problems that human intelligence cannot solve or read minds.

Technological developments show that artificial intelligence will now enter all aspects of life, down to its “capillaries”, becoming an indispensable standard like electricity and the Internet. So what is being done in this field in Turkey?

The Turkey Artificial Intelligence Initiative (TRAI) has been operating since 2017 to raise awareness of artificial intelligence in Turkey and contribute to the development of the ecosystem and cooperation in this field. This ecosystem includes private companies, technology startups working in the field of artificial intelligence, academics, artificial intelligence centers and public institutions such as TÜBİTAK.

TRAI Senior Advisor Roksan Kaspi answered DW Turkish’s questions about what has been done in this area in Turkey and what awaits Turkey in 2024.

Artificial intelligence expert Roksan Kaspi
Artificial intelligence expert Roksan Kaspi Photo: Privat

DW Turkish: How many companies in Turkey are working on artificial intelligence? In what areas do these companies operate?

Roksan Kaspi: There are currently over 320 AI startups in our ecosystem. We regularly update our map of AI startups. They mainly focus on technologies such as machine learning, prediction, data analysis and image processing, which we call machine vision. Natural language processing, chatbots, RPA and autonomous vehicles are among other technologies that startups are focusing on. core products like ChatGPT and Midjourney that have become popular recently. productive artificial intelligence technology Some new initiatives have started to emerge in this area. 10 productive AI startups have been added to our latest map. We expect this number to increase rapidly in the coming years. On the other hand, almost all of the largest top 500 companies have put artificial intelligence on their agenda; They have started adding artificial intelligence technologies to their processes by developing their own technologies or using off-the-shelf solutions. We expect artificial intelligence to be a technology that will be used by all private companies and institutions in a short period of time, a very clear part of our lives and added to every product we use.

Customer analysis, chatbot, robotic processes

What are some examples of innovative apps that have come to the fore this year?

We have many startups working on document reading and recognition, machine vision and image processing. For example, it can be an invoice or an ID; Applications that use artificial intelligence to understand what such documents actually are and who they belong to have now developed considerably.

There are platforms that make it much faster and easier for us to develop machine learning models and significantly increase our predictive success rate. There are many artificial intelligence products focused mainly on marketing. These technologies analyze the customer’s calls or correspondence, understand the level of satisfaction and decide with artificial intelligence what to recommend to the customer next. There are many platforms based on eliminating manual background processes that do the things that humans do much faster by teaching a machine to do the same things through robotic process automation. In generative artificial intelligence, there are some initiatives aimed at generating new patterns or producing new music. There are quite advanced initiatives in chatbot; Of course, with ChatGPT, this business has started to change its model a bit. Now, instead of building their own chatbots, companies are moving to add their own data to ChatGPT and reliably use the bot right there.

Chatbot symbol ChatBGT
Chatbot ChatBGT was among the notable topics this yearPhoto: Andreas Franke / Image Alliance

Looking around the world, what stood out in AI technology this year?

In fact, everyone is talking about productive artificial intelligence. Big language models like ChatGPT and Bard or image generating technologies like DALL-E and Midjourney are on the agenda. Countless new products are released every day for the production of games, audio and video content. Products like ChatGPT or Google’s newly released Gemini seem to stand out among them. This is because these models are multi-mode advanced models. In other words, models that are not focused on a single subject, but are trained with all kinds of data types and have the capacity to produce from all of them. On the other hand, chip technology developed for artificial intelligence has become a very critical issue. There is also a race to manufacture chips that process data faster. That is why we are increasingly moving towards artificial intelligence assistants who can process data faster, process more diverse data and support us in all aspects.

Large language model trained in Turkish

So to what extent are the laws on this issue?

There is a very serious need. This is probably the second biggest trend. Soon “AI law” of the European Union With its adoption, it allowed regulation to move to the next stage, opening the way for all global regulations. On the one hand, companies and major governments continue to determine the ethical principles of artificial intelligence by publishing policy documents. However, such a comprehensive regulation, which the European Union planned to introduce, did not yet exist. That was the first time. Then most likely many states quickly regulations They will further determine.

Are there any studies on this topic in Turkey?

Turkey also has studies on this topic. Work is underway to create an AI Code of Ethics and an AI Policy Document. New AI legislation and regulation is sure to follow. TÜBİTAK has already made a statement on this topic.

On the one hand, there are efforts to develop a major language model trained in Turkish in Turkey. Because all the major language models that have been created so far do not reflect our culture, some patterns of how we use Turkish, and our social perspective, because the data they are trained on is mostly in foreign languages. Or it may contain some prejudices and categorizations that do not exist in our culture. To prevent this, TÜBİTAK is trying to develop models that are trained in Turkish and more suitable for our culture and social perspective, based on the principle of responsible and fair development of artificial intelligence.

TÜBİTAK, Teknokent and R&D support

What is the support for AI companies in Turkey? Are these supports sufficient?

TÜBİTAK calls regularly. The last ecosystem challenge number 1711 has been closed. Through these calls, an expert from the academic world, a technology company and a private company or institution that wants to implement this project present their projects, subject to the establishment of a consortium. While 32 projects applied for the last call, 16 were accepted. Of course, incentives are provided here to implement this project. On the one hand, companies also have various incentives through their own R&D centers or Technoparks.

These incentives must and will certainly increase. The development of models, especially in the field of productive artificial intelligence, requires significant costs both in terms of server capacity and the creation of GPUs and other technological infrastructure. Investments and incentives will therefore have to increase. If we want to become a world-class player, more resources will need to be shifted from the state.

“Awareness needs to be raised”

So where is Türkiye currently in the world when it comes to artificial intelligence?

America and Canada are world leaders in this field. Developed European countries such as Germany, England and France follow slightly behind. We know that intensive studies are also taking place in the countries of the Far East. There is very serious investment going on in China, but it’s a bit of a closed box, we can’t fully follow developments. Of course, it’s hard to keep up with them. Türkiye announced its AI strategy two years ago. Accordingly, Turkey aims to be among the top 20 countries in the ranking of international artificial intelligence indices. There is also a 5 percent contribution to GDP and an employment target of 50,000 people. So actually Türkiye is doing something about it. There are also countries that have not announced a strategy on this issue. It can be said that we are one of the leading countries in the world in these matters. Still, more needs to be done and awareness raised.

Robot unveiled in Bremen
Could artificial intelligence take people’s jobs? Photo: alliance Bastian/Caro/picture

What will be the trends in artificial intelligence technologies in 2024?

Generative AI will continue to evolve in 2024. New initiatives and products will continue to emerge. Current models are already evolving very quickly and will continue to evolve at an even faster pace. Maybe there will be new versions and other capacities that we can’t predict yet. Productive AI technologies are now starting to do things that even humans can’t do. That is why it is the most discussed trend.

Of course, while this technology is evolving, the parallel trend is that companies are now making it applicable in every field by supporting their own products with productive AI products. So what is it? ChatGPT is a chatbot. But every company now dreams of adding its own data to this infrastructure and transferring many tasks in its processes to artificial intelligence. For example, these are technologies that will radically change the customer experience, communication with customers and marketing strategies. Or, employees within a company will now be able to ask AI-based bots about any issue that interests them about their company and how to do which job. In fact, there is such a good assistant, think of him as someone who works with us and knows everything about the company, speaks like a human and understands human language. We no longer need to know technology or be software developers to communicate with artificial intelligence. Thanks to this technology, we are taking our business to another dimension and transforming it.

“The talent gap is a bigger problem in Turkey”

Artificial intelligence will take people’s jobs There are concerns about that too. What is waiting for Turkey at this moment?

Of course, with productive artificial intelligence, fears and anxieties about losing jobs have also come to the fore. On the other hand, completely new areas of business that did not exist before began to emerge. One of them is, for example, the people who write the query ensuring the most suitable output from productive artificial intelligence products, which we call prompt design or engineering. The more correct questions you write, the more successful output you will get. On the one hand, it is expected that there should be ethicists and artificial intelligence managers. Every company will have AI ethics policies and there will need to be managers who oversee and determine them. There will be people who manage the AI ​​processes. There will be positions for people to develop these products and develop these ideas.

On the other hand, there is a serious talent gap in this regard. There are very few people who will provide the technology, infrastructure and software to do this. It is a problem in Turkey and in the world, and it is an even bigger problem in Turkey. Because people who have already reached a certain level in this regard have started to export their skills abroad or are going abroad. Therefore, while job loss concerns are growing, we also have a talent shortage problem. For this, it is necessary to support people with technical and awareness training and increase their talent capacity in this regard. Perhaps departments related to this should be opened at universities. Now there is an incredible resource, there are online courses. There is a need to encourage and guide the new generation into these fields. This is a topic in itself and will be an important topic for 2024.

How can I access DW Turkish without censorship?

Leave a Comment