Generative artificial intelligences (AIs) will facilitate access to knowledge, which will ensure a true meritocracy, says McKinsey global partner Alexander Sukharevsky. For the consultant, this equality of opportunities would resolve, in the future, the concentration of power and money in the hands of a few.
Today, the main AI platforms are under the ownership or influence of the Americans Microsoft and Google.
At McKinsey since 2005, Sukharevsky leads the consulting firm’s AI initiative, Quantum Black, whose goal is to increase the productivity of companies with technology. He was born in Israel and graduated in Computer Engineering from Tel Aviv University and tries to program every day to stay abreast of new technologies.
The executive came to Brazil for business meetings with businessmen and managers and received the report on the 23rd floor of McKinsey headquarters, in the south zone of São Paulo. Quantum Black’s mission is to make artificial intelligence solutions reach day-to-day business —according to him, only 11% of them pass from the pilot stage to the operational stage.
For Sukharevsky, moving from being a data supplier to becoming an exporter of deep learning technologies depends only on Brazilian entrepreneurs.
We’ve had peaks and valleys of enthusiasm for artificial intelligence since the 1950s, when the term was coined by John McCarthy. This generative AI scenario is different from the past? From a technology perspective, this is the first time since 2007 that we’ve seen the birth of new platforms. The first smartphone with access to social networks and the cloud appeared, and suddenly we began to see the emergence of many applications on the market. Right now, we are at a product development plateau.
What I see now especially in terms of product development as well as venture capital is a very strong interest in generative AIs, especially those presented on platforms. Augmented reality, virtual reality and Web 3.0 are also gaining attention.
The generative AI we already see being applied with our customers is still redefining our interaction with the world. In the corporate world, we see co-pilots paid by executives, built for blue-collar workers, to help them do a better job. This is quite different from what we’ve seen in the past.
We see these projections about global GDP growth, increased productivity, but little is said about the risk of strikes and legal challenges coming from new AIs. Could this dehydrate economic momentum? My response as a human being is that AI, like any new technology or anything new in life, has enormous potential but also involves tremendous ethical, cybernetic and humanitarian risks. Before applying it, it is better to know what is being done, in what context it is applied, as we do with any tool. A hammer can build a house or kill someone. McKinsey has an area of knowledge that we call digital trust, for technology in general. We also develop AI for the good of humanity and we still cannot quantify how much this technology can help create solutions to old ills and new advances.
What is the competitive advantage of building your own AI model versus one built by OpenAI [criadora do ChatGPT] or by Google in building a platform? What I try to do is work with creative people to understand what’s missing to put these plans and tools into practice. You can have multiple base templates. Language models change based on how the organization takes its data and asks the model questions. These are not just generic questions, but those rooted in the reality of the company. The second point has to do with attracting the best talent available. It’s not just a technical issue, it’s more of a business issue. After that step, we can discuss ethics. It’s not just a discussion of which engine to use, but the whole car.
Among the 63 sectors of the economy mentioned in the McKinsey study, are there companies from Brazil? In many. That’s why I say I’m inspired by the level of talent in Brazil and the level of innovation. Everything that has to do with marketing and sales, Brazilian creativity in terms of advertising. Brazilian agencies, for example, are recognized with awards around the world. This will reach the advertising market, customer service. The other thing is the creation of a virtual brand identity, like a company chatbot. You will be able to call and speak to the bank officially. You start talking to your supplier, and those conversations can help provide better answers in the future. This will improve the way companies serve customers.
Behind the new AIs, there is a lot of work of image description, translation and transcription, done by workers who do not receive proper instructions. To have safer systems, it is necessary to increase investment in the treatment of data? This is indeed recommended. What we’ve seen more recently are more and more tools that allow us to do high quality work at affordable costs. It is also possible to allow data annotators to use copilots. But without the right data, all the rest of the work falls apart.
Mr. Do you foresee that you will continue with your current 5,000 employees producing more and more? Or that you will have 4,000 doing the work of the current 5,000? Our goal is to achieve the best technology mindset at McKinsey, but also to continue to grow the company. As this is the age of creativity, I would love to have even more employees with me. There will be more productivity and it saves time. But we are always focused on inclusive and sustained growth. As I assess that we’re going to be able to generate a lot of growth and impact in the public and private sectors, I would probably need more people.
Mr. commented on the ethical use of artificial intelligence and one of the discussions is about a possible increase in inequality, as is the case in Brazil. How to deal with it? With the numbers we have today, I strongly believe that open source models will win the dispute over proprietary models [como é o ChatGPT]. With a strong community, open source can deliver better results. This current age of creativity gives us more knowledge. People now have tools at hand to create solutions. Before, they didn’t have that same access. It will be possible to create many new products. Think about education, it will be possible to take the best knowledge to different places. It’s a cause that fits in your pocket, although today you can’t afford it. It will help in personal development, in communication. With these tools, we can narrow the gap between people and create opportunities for almost everyone, which can allow for an era of real meritocracy.
Is it possible that Brazil will end up as a data provider and not be able to develop its own AI technologies? It will depend on Brazilian entrepreneurs to make the country move away from being a supplier and becoming a producer. Two factors that can help: the size of Brazil, which is one of the largest economies in the world, and investments in generating AI in Portuguese. This is a relevant challenge for the internal market. From an internal perspective, it is essential to promote the country’s entrepreneurship, really creative people. It’s a country that, based on its history and what’s been happening over the last 20, 30 years, is constantly reinventing itself and finding new ways of doing business. Emerging economies have cycles, with ups and downs, but Brazilian entrepreneurs and Brazilian companies are always thriving and creating excellent business models out of difficulties. This will be driven by generative AIs.
Alexander Sukharevsky, 44, leads McKinsey’s AI initiative Quantum Black, which aims to make companies more productive with technology. He was born in Israel and graduated in Computer Engineering from Tel Aviv University.