It’s morphing out of a tool for productivity into a technology that could shape scientific research, cybersecurity, medicine and other increasingly important domains. Yet, as the scale of AI systems continues to grow, so do fears over how they might be misused.
Microsoft co-founder Bill Gates has issued a chilling warning about the risks of high-powered AI. Speaking to NBC’s Meet the Press, Gates said that AI was “powerful enough” that it could “allow situations where a billion people could die”. His comments have spurred a new line of discussion on the safety of AI and the responsibilities of tech companies.
Significantly, Gates was not warning that we are on course for a billion deaths from AI. Instead, he was issuing a stark warning about what could happen should more powerful AI tools be used by malicious actors.
What Did Bill Gates Say About The Risks Of AI?
Gates’ worry is about the use of AI systems by bad actors. In his words, “AI could be leveraged by bad actors as a force multiplier, providing access to tech and information that was once confined to big government and big tech.”
This complicates the conversation about AI safety. It’s no longer about a hypothetical where an AI decides to turn on the human race. It’s about how ever more capable systems might assist humans in performing destructive tasks more effectively.
Gates has thus suggested that governments must be more active in determining what types of protections and monitoring should be necessary as AI develops.
AI and the Growing Biological Risk
The largest areas of concern include: Biology.
Today’s AI models can analyze vast troves of scientific data to support researchers working on areas such as drug discovery, genetics and molecular biology. These tools could be instrumental in speeding up drug, vaccine and therapeutic development.
Yet the same abilities could lead to what experts call a dual-use problem. A technology intended for legitimate scientific study might also have the potential for nefarious exploitation.
What Gates worries about is that AI might decrease some of the technical hurdles that currently exist for engineering a harmful biological agent. And a biological attack, he said, might have consequences far beyond the individuals who engineered it.
Cyberattacks Could Become More Sophisticated
2.9.11 Cybersecurity There is a discussion around cybersecurity in the context of risks from AI.
AI could also be used to automate some tasks, analyse large volumes of data and expose vulnerabilities in digital systems. If it is used by the wrong people, it could be exploited to attack vital infrastructure.
Types of Organisations At Risk Organisations that might suffer severe losses due to highly advanced cyberattacks include hospitals, financial institutions, energy providers, and government agencies.
This concern is arguably especially relevant since AI may even be used to help perform some of these complex digital activities with less specialised expertise.
Gates Sees Both Opportunity and Risk
But it isn’t all bad. Despite the warnings, Gates himself has said that he sees the potential for AI to revolutionise how we care for the sick, learn in school and farm.
His recent public remarks rest on a longer argument that AI might emerge as one of the most transformative technologies for human flourishing, but its overall benefit will be contingent on its safety.
Gates also sits on the board of another Big Tech foundation, Microsoft, which has been pushing for the use of AI in a wide range of sectors, including healthcare, agriculture and education. Mark Zuckerberg, Facebook CEO. (Photo by David Ramos/Reuters/Corbis) AGf.
The Challenge ofAI Regulation Is Growing Why?
Gates has said that AI safety shouldn’t rely solely on pledges from tech companies. He has said that governments and lawmakers must get involved in setting up safeguards and monitoring.
The issue arises as governments and AI companies wrangle over how AI systems should be evaluated, overseen and, ultimately, regulated. The difficulty is devising protections that counter severe dangers but do not hold back harmless innovation.
This debate could grow more significant for business, investors and technology users as AI permeates more industries.
The Bigger AI Question
The billion-death warning from Mr. Gates is not really a prediction but about a worst case. Technology is rapidly advancing and AI is getting better at doing more complex things.
This is a strange paradox: the very same technology that can be used to advance the discovery of medicines, improve the provision of education, and bolster our productivity, may, at the same time, be used to enhance the ability of malicious humans.
With the continued development of AI, the argument has now moved from the question ‘What can AI do?’ to the question ‘How should we regulate and use it?’.
For governments, tech firms and society at large, that could be the defining challenge of the AI age.
