Transparency
and efficiency are crucial to regulating artificial intelligence. The Russian
Federal Service for Environmental, Technological, and Nuclear Supervision is
already taking on the role of a certifier of solutions that use AI
technologies. The creation of testing grounds for AI is becoming a critical
phase on the path to innovation. The lack of clear regulatory mechanisms
further complicates the introduction and testing of new AI systems. However,
granting small and medium businesses access to existing and new testing grounds
will eliminate many of these barriers. Introducing such testing grounds will
serve as a catalyst for the further development and creation of new
technologies. These steps will not only ensure the safer and more efficient use
of AI, but will also provide a broad range of opportunities for innovation that
could radically change future industries. This was the opinion reached by
participants in the session ‘Testing Grounds for Artificial Intelligence at
Russian Energy Week 2024.
KEY CONCLUSIONS
The regulation of artificial intelligence
should be transparent and effective
“A large
company creates a testing ground. It says, I created a testing ground and I
test everything. First, it’s an analogue of various scientific and technical
centres that previously handled basic equipment of different kinds. It’s a sort
of restrictive thing for procurement procedures. And businesses are often
unhappy with this. A filter will appear here at the entry point regarding art
and intelligence. Everyone works according to their own standards. I always ask
the question: how much does a smart fan cost and how much time and money will
you take from me so that I can receive a piece of paper with a blue seal, and
this piece of paper is recognized on the market. When the piece of paper is
recognized on the market is a question from the regulator to us and other
agencies. It’s mutual recognition. But how should we deal with the fact that
this is a business entity? As an official, I can scare a business entity, but
it will hang up on me. So, I scared it and that’s great,” Eduard Sheremetsev, Deputy
Minister of Energy of the Russian Federation.
Experiments in the digital economy have helped
bypass more than 150 regulatory barriers, which accelerates the introduction of
new technologies
“At
present, thanks to the current experiments, we are bypassing 30 regulatory acts
and over 150 barriers in these acts. The participants include more than 190
companies and government agencies. Over 100,000 citizens took part in this when
they were seeking telemedicine services, were robotaxi passengers, or played
other roles,” Vladimir Voloshin, Director of the Digital Economy
Development Department, Ministry of Economic Development of the Russian
Federation.
The second cognitive revolution is affecting
all sectors of the economy, including energy and manufacturing, which will lead
to major changes in industry and technology
“Today, we
can say that the modern world is moving, as I would put it, towards a new tectonic
shift, which we can call the second cognitive revolution. This is precisely
what the second cognitive revolution is – the symbiosis of machine and man. And
soon it will be extremely difficult to recognize what is real and what is
virtual [...] The second cognitive revolution, of course, concerns not only
energy, but everything, all sectors of the economy, primarily manufacturing,” Nenad
Popovic, Minister Without Portfolio of the Serbian Government.
AI needs to be developed to improve and create
new technologies
“Artificial
intelligence has traditionally developed in virtual environments. But there are
no effects for industry there. For us, the effects occur when we start using
artificial intelligence in the real world. And this is where the problems really
begin, because the real world is quite different from the virtual one. Even at
the fundamental level, if we need to explain to artificial intelligence what an
apple is, we will explain that it is green and round, but it’s just a virtual
object. In order for it to truly understand what an apple is, we need to
explain that it is hard, but if you press hard, it stops being hard. And it’s
the same with other objects too,” Alexey Shpilman, ITMO National Research University
Professor.
PROBLEMS
Insufficient regulation of artificial
intelligence
“The
technologies that spring to life on an inventor’s desk appear in series
production in two to three years. We aren’t that different from those who make
airplanes. They also test aircraft construction for years. We are just as
responsible in this regard,” Eduard Sheremetsev, Deputy Minister of Energy of
the Russian Federation.
Lack of a testing ground for artificial
intelligence
“Today,
there are already examples of testing grounds for artificial intelligence and
digital technologies, for example, in nuclear energy. Rostekhnadzor [the
Russian Federal Service for Environmental, Technological, and Nuclear
Supervision], as a regulator, also uses the testing grounds of companies for
the scientific and technical support of our subordinate enterprises, which
create an idea of the models that are capable of properly representing complex
technical objects,” Stanislav Korchivoy, Deputy Head, Federal Service
for Environmental, Technological and Nuclear Supervision.
No clear mechanisms or legal regimes for testing
artificial intelligence
“There are
three things that are essential. It’s important for it to be clear and simple.
It’s important for it to be of high quality. It’s important for it to have good
speed [...] It’s important that we truly shorten and not lengthen the
bureaucratic path. It’s essential that it’s clear where people need to apply so
that there is a one-stop shop for a system of testing grounds or one large
testing ground. The quality must be high, so that qualified people who
understand how to conduct proper tests are present at the testing grounds. It
must be simple. It must be of high quality. It must have good speed,” Alexey
Shpilman, ITMO National Research University Professor.
SOLUTIONS
Provide small and medium businesses with access
to testing grounds
“For small
and medium businesses that operate hazardous facilities, the government could
offer serious support in terms of standardization and providing the opportunity
to conduct research at such testing grounds,” Stanislav Korchivoy, Deputy
Head, Federal Service for Environmental, Technological and Nuclear Supervision.
Create testing grounds for artificial
intelligence
“As a
state, we must create such testing grounds so that they have potential. That’s
why I believe in the driving force of Russian, Serbian, European, and global
startups, which are creating the next platform for the development of
artificial intelligence related to the robotization of industry. It’s crucial
to create our own system. Let’s create our own system to standardize
statistical monitoring. We will understand what is happening with AI and
robotization. And targeted research programmes for science and innovation
ecosystems,” Nenad Popovic, Minister Without Portfolio of the Serbian
Government.
“We are
always talking about digital sovereignty and technological sovereignty, but
there will not be any technological sovereignty or digital sovereignty without
sovereignty in robotics. We must make production in all sectors of the economy
faster, more automated, and more series-produced. And I see great opportunities
for our bilateral cooperation here,” Nenad Popovic, Minister Without Portfolio
of the Serbian Government.
“To do
this, we all need to get together. Not just large companies and small
companies, but also regulators and the federal executive authorities, so that
we can all work together to develop a unified system, including in terms of
legal regimes. Perhaps experimental legal regimes are needed. These are all
open questions […] It’s important for this to be clear and simple,” Alexey
Shpilman, ITMO National Research University Professor.
* This is a
translation of material that was originally generated in Russian using
artificial intelligence.
For more
information, visit the Roscongress Foundation’s Information and Analytical
System at roscongress.org.
