ASTANA – The first Connected international conference, held in Astana on Oct. 18, brought together experts to explore the future of the economy, climate change, and data security, with a focus on the evolving role of artificial intelligence (AI) in shaping these areas.
The conference, focusing on conceptualizing the vision of the future, brought together 1,500 delegates from over 20 countries.
“We have invited researchers, futurists, leaders in innovation and technology, and public and government figures to the conference. Together with them, we hope to achieve the main goal of the Connected conference — to find a path to global harmony in the future. To do this, it is necessary to create a vision of the future world, including our country, over the next 20-30 years,” said Serik Tolukpayev, entrepreneur, investor, and founder of the Aitas agro-industrial holding, general sponsor of Connected.
During the opening ceremony, Deputy Prime Minister of Kazakhstan Tamara Duisenova read a welcoming speech from President Kassym-Jomart Tokayev.
“Kazakhstan, in cooperation with leading foreign universities, is actively working to create an academic hub. We are forming a solid foundation for increasing the scientific and innovative potential of the country, and the Connected-2024 conference will give impetus to achieving this goal. I am confident that today’s event will contribute to strengthening international ties between scientists and experts, as well as developing new approaches to education and human capital development,” said Tokayev.
How is AI transforming the future landscape?
The plenary session titled An Informed View of the Future: A Worldview Approach opened the conference, gathering futurists, historians and AI experts from Switzerland, South Korea, the United Kingdom and the United States.
Can AI save the world or help combat climate change, and how to do it while maintaining the ethics, sustainability and security of AI – those were the questions asked to plenary session speakers by the moderator, journalist and former head of CNN International and Al Jazeera English, Riz Khan.
According to Gerd Leonhard, futurist, humanist and author of “Technology vs. Humanity” from Switzerland, it is important to be cautious of AI technology’s side effects as it undergoes rapid growth among consumers and economies.
“Artificial intelligence as a business tool is basically just smart software. So that’s probably a good thing, but to seek to build a machine that is more powerful than the human, like OpenAI wants to do, is basically like saying ‘I’m going to run the fossil fuel economy and not pay any attention to the side effects.’ So we need to think about the side effects today. We need to think about how we mitigate the side effects. Because unlike climate change, which we can solve, we probably cannot solve general intelligence once it’s here,” said Leonhard.
The role of AI should solely be limited to being a tool for building a better future, according to him.
“We don’t need machines that have consciousness. We need machines that are competent, and I think this is the priority, that we use them for stuff that fixes problems. That’s not to use a machine to be like a human. The problem is when you have a machine that is kind of like a human, then we become the second most intelligent species, and that is a very bad place to be,” said Leonhard.
Regarding AI impacting job security, he noted that humans are far more complex than mere computing mechanisms. Therefore, the focus should be on areas where AI cannot compete or replicate human capabilities.
“We’re moving in a sort of a future beyond the simple knowledge. So machines can have knowledge, but they don’t have deep knowledge, or quiet knowledge, or human agency. So, the future of our work is human-only work, and that is creativity, design, imagination, negotiation, intuition. And this is what we have to teach our kids,” he said.
AI and climate change
Kay Firth‐Butterfield, Good Tech Advisory CEO and former head of AI and member of the Executive Committee at World Economic Forum, challenged the narrative that AI can help to combat climate change, as the technology relies on large data centers that demand substantial amounts of electricity.
“Every time we ask an LLM (Large Language Model) a question, we are not only using a lot of electricity, powered at the moment by fossil fuels, but it is also using a quarter of a liter of water. So I encourage you, when you’re asking these LLMs questions, to ask them sensible and important questions because the climate impact of you asking those questions is enormous. We’re looking at 3.7% of global greenhouse emissions for AI. That is more than the aviation sector,” said Firth‐Butterfield.
Rae Kwon Chung, Nobel laureate in green economy and director of Ban Ki‐moon Foundation for a Better World, advocates for a shift from traditional economic growth models to a sustainable model, which promotes economic progress while reducing environmental impacts. In the meantime, the free market system should be more quality-focused.
“The liberal democracy and the free market has been excellent and brilliant in increasing the quantity of production and increasing the freedom of the number of the voters around the world for free election. But increasing quantity was not enough to improve the quality, so we have to enter into a new phase of development where the quality matters, not just the quantity, but the market function,” said Chung.
“We have such an imbalance between the market, the people and the planet, and at the expense of a huge progress in quantity,” he added.
However, transitioning from a free market to a sustainable market cannot happen overnight; an intermediate step is necessary.
“I’m arguing that we can have a middle step, middle ground, which is a voluntary market based on the consumers who are ready and aware about climate issues,” said Chung.
Blaming governments for inaction on climate change is not enough, according to him.
“The Paris Climate Agreement is a top-down by the government, and we have to design a way that the consumers can share responsibility. We don’t have the system enough. We have to encourage the consumer because without changing the consumer’s behavior, it is very difficult to have a different result,” said Chung.
“Very interestingly, even the business is interested in coming up with a carbon-free branding. If their brands are carbon-free, they believe they will have a stronger marketing power,” he added.
AI and data accuracy
Another problem with AI is that the data in large language models is often based on outdated or biased information.
“The majority of data in the large language model is from the internet. Whose thought processes have populated most of the internet? – White men. Women didn’t really start creating very much that is part of the internet until maybe the 1950s, maybe the 1960s. Women in the West were not allowed to participate in medical studies until 1965. So, if you’re looking for data about women’s health from an LLM, you’re not going to get as much about women as you would from men. If you want to use an LLM to think about curing heart disease, the data is primarily drawn from white men living in America over 1955. For persons of color, of course, the data is even more poor on the internet,” said Firth‐Butterfield.
Leonhard echoes her points by saying that AI is reductionist, which is “they only see the obvious.”
“For example, if you look into hiring a chief data scientist, the system would not show the ad to any women because women aren’t data scientists, according to the machine. It’s the opposite of what we want,” he said.
“We cannot expect the machine to understand what is coming. The machine understands what was before, and so it will recreate what was before, apart from, of course, not knowing all the languages of the world. That’s another big problem. It’s mostly English,” added Leonhard.
Among other topics covered by the conference are technology and human synergy, sustainable development, climate future, and global competitiveness of higher education and science.