Vision vs ChatGPT: two answers compared

 

The first attempt to regulate Artificial Intelligence dates back to 1950, when Isaac Asimov published the short-story about brain-equipped robots.

“A robot may not injure a human being or, through inaction, allow a human being to come to harm. A robot must obey the orders given it by human beings except where such orders would conflict with the First Law. A robot must protect its own existence as long as such protection does not conflict with the First or Second Law”. 

After 73 years, it seems we are truly on the threshold of a revolution that could lead us into a world whose contours are not even imaginable. The question is the same that Asimov tried to answer: is it possible to regulate Artificial Intelligence? Is the European approach effective, considering the European Union is cut off from a battle where Americans and Chinese lead the way? In this paper, Vision's team attempts to give an answer to these fundamental questions. The final section of the paper presents a response provided by CHAT GPT - an Artificial Intelligence system itself. 

 

LEVEL OF AI DEVELOPMENT IN THE TOP 10 COUNTRIES 

(WEIGHTED AVERAGE BETWEEN THE LEVELS OF IMPLEMENTATION, INNOVATION AND INVESTMENTS)

SOURCE: VISION’S DATA BASED ON THE COUNTRIES TOTAL RANKING POSITION IN THE GLOBAL AI INDEX RANKING (by TORTOIS)

 

Chat-GPT, the app developed by Open AI – a no profit organization, that, however, is funded by Microsoft to take a share of the most profitable market, the digital marketing one – has already become the most quickly adopted product in history: 100 million users in two months since its launch.

The chat, providing accurate and immediate answers to complex questions, is likely to devour the space occupied by hundreds of “intellectual works” that, from universities to courts, try to summarize what other say. The competition to control search engines, Internet’s nervous system, is even more crucial: a digital assistant which searches useful insights on the network organizing them, crowds out the same method that has been used in the past two decades by the business, Alphabet – which has become Internet’s nervous system. And, yet, Chat-GPT is not alone. Alphabet replies with Bard, Baidu – the Chinese Google – with Ernie. Beside Internet giants, very talented start-ups are also moving: NEEVA, You.corn, C3.ai.

The European absence is, for a change, concerning. The European Union is not even taking part to the match for global digital leadership – differently from India or the United Kingdom. Instead, the European Union leads the way of producing laws on the digital world, (Vision Think tank counted ten laws, with 563 articles, in the last eight years).

THE EUROPEAN PRODUCTION ON DIGITAL  2016 - 2022)

Source: European Commission and European Parliament

This attempt encounters three conceptual problems, even more evident when you try to control AI.

Firstly, defining a phenomenon discussed for ages by scientists and philosophers is a tough task. The latest proposal by the Council openly tries to set a more precise definition (so to avoid to include all kinds of software) but the reference to a “system that suggests on the basis of some sort of learning” can still be extended to almost everything - including movie libraries as Netflix.

Secondly, it is difficult to regulate a phenomenon so fast that after the latest common position of the Council was adopted (November 2022) some extreme changes such as Chat-GPT launch occurred.

Finally, we refer to global processes that, by definition, need global government tools. If we limit ourselves to regulate them from European boundaries, we risk to get hurt twice. We would not only avoid some of the most feared implications: preventing  the use of biometric identification system in public places in Europe can even accelerate such experimentations in unscrupulous countries, waiting to spy on us through satellites. We would also lose potential benefits: claiming that self-driving cars are monitored by a person means to definitely abandon one of the most interesting possibilities for the car industry to escape its own obsolescence.

The alternative to such regulatory bulimia, which seems to be a conditioned reflex of the European Commission, may be a pragmatic approach as the English one – with the Competition and Market Authorities: intervening on single, but big implications of the technological progress, balancing, on a case by case basis, risks and opportunities.

The European Union can no longer be based on the assumption that it should defend itself from phenomena that we are forced to experience. 

Even Asimov was aware that its three Laws led to disturbing paradoxes. And yet, we have to face the future with the controlled ambition of those who know they do not have alternatives. That was after all the ambition of the European navigators that changed history daring to explore completely new worlds. 

 

---

CHAT GPT'S ANSWER

Is Artificial Intelligence can be regulated by governments? So that benefits can be maximized and risks be reduced? How would you suggest to do so?

Yes, governments can regulate artificial intelligence (AI) in order to maximize benefits and reduce risks. In fact, many experts argue that AI should be regulated in order to ensure that it is developed and used in a way that is safe, ethical, and beneficial for society as a whole.

Here are some ways that governments can regulate AI:

1. Creating regulatory frameworks: Governments can create regulatory frameworks that establish guidelines for the development and use of AI, as well as mechanisms for enforcement. These frameworks could include rules and standards around issues such as data privacy, security, bias, and accountability.
2. Developing ethical guidelines: Governments can work with experts and stakeholders to develop ethical guidelines for the development and use of AI. These guidelines could help ensure that AI is developed and used in a way that is consistent with values such as fairness, transparency, and accountability.
3. Investing in research and development: Governments can invest in research and development related to AI, in order to help ensure that it is developed in a way that maximizes its benefits and minimizes its risks. This could include funding for research on AI safety and security, as well as for the development of new technologies and applications.
4. Supporting education and training: Governments can support education and training programs that help individuals develop the skills and knowledge needed to work with AI in a safe and ethical manner. This could include funding for programs in areas such as computer science, data science, and ethics.
5. Engaging with stakeholders: Governments can engage with stakeholders from industry, academia, civil society, and other sectors to ensure that the regulatory framework for AI is responsive to the needs and concerns of all stakeholders. This could include mechanisms for consultation, collaboration, and public participation.

Overall, regulating AI will require a multifaceted approach that involves a range of stakeholders and addresses a range of issues. By doing so, governments can help ensure that AI is developed and used in a way that maximizes its benefits and minimizes its risks

Thanks a lot. Now one more specific question. Which are the cons and the pros of the European Union (European Council in fact)'s proposal to regulate Artificial Intelligence?

The European Union's (EU) proposal to regulate Artificial Intelligence (AI) has both pros and cons. Here are some of the most important ones:

Pros:

1. Promotes ethical AI: The EU's proposal promotes the development of ethical AI that is safe, transparent, and respects fundamental rights. This could help build public trust in AI and ensure that it is developed and used in a way that benefits society as a whole.
2. Enhances accountability: The proposal includes provisions that would require AI developers to document their systems and provide information about their decision-making processes. This could enhance accountability and help prevent the development of AI that is biased or discriminatory.
3. Protects fundamental rights: The proposal includes provisions that would prohibit the use of AI for certain purposes, such as biometric surveillance, that could infringe on fundamental rights such as privacy and data protection.
4. Streamlines regulation: The proposal would establish a single set of rules for AI across the EU, which could help streamline regulation and reduce the burden on businesses and other organizations operating in multiple member states.

Cons:

1. Stifles innovation: Some critics argue that the proposal could stifle innovation by imposing onerous regulations on AI developers and limiting the development of new AI applications.
2. Excludes smaller businesses: The proposal could disproportionately affect smaller businesses and startups that may not have the resources to comply with the new regulations. This could make it more difficult for these businesses to compete with larger, more established companies.
3. Difficult to implement: The proposal is complex and could be difficult to implement, particularly given the fast pace of technological change. It could also be difficult to enforce, particularly given the cross-border nature of many AI applications.
4. Potential for regulatory fragmentation: The EU's proposal could lead to regulatory fragmentation if other countries or regions develop their own AI regulations that differ from those in the EU. This could create additional compliance burdens for businesses operating in multiple jurisdictions.

Overall, the EU's proposal to regulate AI has the potential to provide important benefits, such as promoting ethical AI and protecting fundamental rights. However, it also poses some challenges, such as the potential to stifle innovation and the difficulty of implementation and enforcement. It will be important for policymakers to strike the right balance between regulation and innovation in order to maximize the benefits of AI while minimizing its risks.

------------------------------------------------------

^1: The paper has been an input to a column published by Francesco Grillo on the Italian newspaper Il Sole24Ore.