David Yang on the benefits of AI for humanity. Artificial intelligence is entering our lives and affecting its quality. But can the creation of human hands think? What to expect from smart machines in the near future, we talked with David Yang, Chairman of the Board of Directors of ABBYY Group of Companies.
David, why is artificial intelligence important for humanity?
Last year I gave a report on the consciousness of artificial intelligence. At a conference with leading world scientists in this field, AI issues were discussed at the intersection of science and philosophy. So, discoveries in the field of artificial intelligence are epoch-making. Understanding how large linguistic models work, we begin to approach the understanding of what consciousness, thinking, freedom, and will are. Even the Dalai Lama this year did not deny the possibility of consciousness in a computer, whereas he used to deny it.
Will there be practical benefits from AI?
I am a techno-optimist and believe that all historical technological revolutions have led to increased labor productivity and, as a result, improved quality of life. This was the case with the invention of engines, electricity, and all other major discoveries. Of course, there were risk areas at the moment, but humanity must take a conscious position in this sense. We understand that the engineering of fifth-generation self-driving cars threatens millions of driver jobs. But this does not stop humanity from investing in the self-driving car industry because billions of people will gain increased comfort and safety. We just need to learn how to retrain those people who may lose their jobs due to new technologies.
If we talk, for example, about medicine, what revolutionary can be in this area? AI in medicine includes personalized medicine (for example, developing individual drugs that will work strictly on specific patient proteins), precision early diagnosis, remote surgery, and much more. Artificial intelligence, thanks to its generative and predictive properties, can reduce the number of experiments needed to find appropriate formulas by many orders of magnitude. I do not exclude the moment when big data will allow us to cure diseases in ways we cannot even imagine. Drink such a solution, massage such a point on the body, smell this scent, and listen to this music – and the incurable disease is gone. And this will no longer be shamanism, but evidence-based medicine, interestingly, we will not fully understand the mechanism of healing.
Is AI smarter than humans?
In some sense, it has been smarter for a long time. Suppose there is a task to write a program in Python of a certain number of lines, and the program will do this without errors. It would seem, what’s so special about this? In principle, a college graduate or an engineer with two years of experience can also write a program. The amazing power of modern generative models is that they write it “in their sleep,” meaning they do not reason, they dream like poets, their lines are born, they do not think about how to create the next one, they act like artists or composers. Artificial intelligence acts unconsciously, like in a dream. Programmers cannot write a thousand-line program in their sleep.
The next leap will occur when the ability for conscious logical reasoning is added to this absolutely unconscious state. Imagine a brilliant child who cannot consciously speak yet but at the age of two can write thousand-line programs with closed eyes without errors, and they work. And if you whisper in his ear while he sleeps: now change all these variables to these and make it blue. And he does it all. Imagine this brilliant child grows up to 25-30 years old, and he, possessing such supernatural abilities, starts to reason.
Can we teach this brilliant child concepts of morality, conscience, and kindness?
We must do it. A huge number of commissions and working groups are currently engaged in this because a new race of non-biological companions is emerging on our earth. “I predict that in a few years, there will be special regulatory bodies that will give a maturity passport to some of these models that have passed certain tests for compatibility with human values.” In such a case, models will receive part of human rights, and everything is heading towards these non-biological companions having a significant impact on society.
Not everyone may like this. Of course, take the problem of migrants in the US, Europe, and other developed countries. Part of society supports migrant rights, part builds 6-meter fences against migrants. Both have their arguments. But stopping the migration process is impossible, so the truth is in the middle, we need to find ways to legalize migrants and adapt them to the norms accepted in this society. The same will happen with AI agents, digital employees, and non-biological companions. It is impossible to stop the process, but finding mutually beneficial living norms is our duty.
I am sure there will be people who will take to the streets and protest against AI. Opposing them will be people with arguments about how many lives non-biological doctors and drivers have saved. And it will be a new war between the “North and South.” Somewhere in the middle, these two sides will converge, meaning there will be specially regulatory bodies that will test these models for compliance with our ethical norms and values.
David, do you consider the machine uprising a utopia?
It seems to me that this is more of a scary story from science fiction. It will happen a little differently than we imagine. Indeed, domestic humanoid robots will have a high degree of independence, and we will need to learn how to live with them. Scientists predict that by the forties, there will be about a billion of them on planet Earth, they will really be family members. But they are unlikely to have the ability to unite and negotiate; machines will not compete with people, but people, as a very complex and developed biological species, will continue serious intraspecific competition, and different smart machines will be part of such a struggle.
For what kind of businesses, in your opinion, will artificial intelligence become necessary? First, I would name the area we call digital employees. This is a special type of intelligent agent that can communicate with the outside world using the same interfaces as human employees. This includes phone, email, messengers, CRM systems, ERP systems, and physical presence in the office – everything related to sales, initial communication with clients. These are entry-level positions with a huge number of vacancies in the US. About 11 million vacant low-paid jobs. People do not want to take positions paying $17-22 an hour in the US, causing both small and medium and large businesses to lose colossal revenue.
We cannot reach the nearby SPA salon, the nearby store, 40% of incoming calls go unanswered, businesses lose their revenue, and we cannot place our orders as consumers. The problem is that this is not a creative job area, it is impossible to pay more for it, it is economically unprofitable, and people take these positions only for a few months and then leave. Today, opportunities have emerged to fill this niche with digital employees: AI receptionist, AI concierge, AI hostess, sales support. They can work around the clock, remember the names of our dogs, cats, and our habits, and as a result, the quality of service we receive improves.
Where else can an AI assistant find an application?
There are specifics for each industry. For example, in law, assistants can perform tasks like checking contracts, incompatibility, contradictions, highlighting facts; for financiers, some forecasts, some document submissions, and simple reporting decisions. Creative people also have a vast field for activity: for example, post-editing has changed a lot, I won’t even talk about translators. Even in areas like design and architecture, huge opportunities have emerged to reduce less creative work.
Machines currently operate at the level of college graduates, meaning a person with ten years of experience, for example, a talented journalist, a professional in their field, will do the job better. But soon, this may all change. In general, everything has changed so radically over the past two years that it is difficult to say what will happen in a couple of years, we will live and see.
If you try to build a business competently with AI now, can you overtake enterprises with many years of experience? From scratch – and to the top?
Do you mean, can a business owner significantly increase their economic performance by using artificial intelligence? Of course. For individual tech companies, these technologies provide multiple opportunities. Even for the “real economy,” there are areas with the potential to significantly improve performance. At the Digital Worker Forum conference, which we organized in Sunnyvale in 2023, a company spoke that uses weed recognition in video streams in real-time and a laser to precisely destroy them without chemicals, significantly increasing productivity.
Overall, due to artificial intelligence, labor productivity per capita worldwide will increase significantly. As a result, the quality of life will improve, for example, our workweek will average from 5 to 3 days. In our free time, we will travel, engage in creativity, new companies, new projects, learning, and self-education, generating new industries. This is how it all works in the economy, just as it happened in past technological revolutions, so it will happen now.
We see what is happening in the world now. Will artificial intelligence help humanity stop wars?
If we delve a little deeper into how evolutionary selection occurred in various social groups, it becomes clear that interspecies and intraspecies competition within the boundaries of “us-them” is inevitable. It is tragic for us to realize this from a humanitarian perspective, as we, Homo sapiens, hoped that evolution would eventually take forms other than destruction. But 40,000 years before our era and the last two thousand years of our era’s history have been marked by a huge number of wars. Less than 70 years after the end of World War II, we again see a serious escalation. I am afraid that artificial intelligence will not be able to help humanity find more humane ways of competition.
newo.ai – Digital Employees. The revolutionary project by David Yang.
Digital employees are cloud-based software complexes based on LLM. Unlike regular AI agents, they have key features: 1) the ability to communicate with customers and employees of the organization through regular human communication channels (phone, email, messengers, etc.), 2) the ability to perform complex multi-step actions (reservations in restaurants, placing orders by phone with CRM/ERP entry, etc.) receiving tasks from the manager in the form of instructions in human language, 3) presence in the physical world in the form of AI kiosks or humanoid robots. They can “work” as receptionists, concierges, salespeople, and significantly increase business efficiency.
Digital employees provide services, process customer requests, use corporate knowledge bases and information systems (CRM, ERP), perform intellectual tasks, comply with corporate policies, and achieve set KPIs.
(When you register using the unique link above, you will receive free basic implementation of an AI receptionist/hostess and a $300 credit for platform use and traffic payment)
