The Importance of Ethics and Governance for Artificial Intelligence

The Promise, Possibilities and Potential of Artificial Intelligence

The utilization of artificial intelligence has brought huge benefits to current and future societies, such as improving supply chain efficiency, promoting Smart automation and robotics to support our daily lives and provide seamless customs clearance at borders. Artificial intelligence technology has advanced so much that it is rapidly approaching, and in some cases even surpassing, human intelligence. Just as human intelligence needs to be properly harnessed and coordinated within a human society with good ethics and governance, we need to do the same with artificial intelligence.

The Dark Side of Artificial Intelligence

There are some things to keep in mind. If you’ve heard of GPT-4chan, you’ve also heard of its problems. The GPT-4chan model was developed by a Youtuber in the artificial intelligence (AI) community and is based on the /pol/ (politically incorrect) section of 4chan, a controversial forum. 4chan's /pol/ (abbreviation for "politically incorrect") section is home to hate speech, conspiracy theories and far-right extremism, and is 4chan's most active section. The results of it? The result is an artificial intelligence that spews hate speech.

According to reports, Yannic Kilcher, a well-known YouTube deep learning blogger, used 134.5 million posts full of verbal violence to train an AI model called "GPT-4chan". Through training, it learned how to Comments were posted on the site and more than 15,000 posts filled with negative content were published in less than a day.

Although this may seem disturbing, the foundations of artificial intelligence may inadvertently encourage this practice, making it easy for people to use the technology for nefarious purposes. This is very evident as the community increasingly embraces open source development. This no longer limits the development of AI applications to a privileged few companies, but opens them up to everyone, including bad actors.

Should we be surprised? Celebrities including Bill Gates, Elon Musk and Jeff Bezos have expressed concerns and warned about the potential dangers of AI, particularly around its use in weapons systems and job losses .

However, most of what we experience today involves artificial intelligence in some way, such as so-called personalized recommendation ads on social media that are targeted specifically at us, and our favorite streaming platforms based on our Previous choices and habits recommend new content to us. If we really want to turn the potential of AI into a reality, there are some issues that must be solved first.

Growing Concerns

Artificial intelligence systems have grown exponentially in recent years. It has produced many benefits, but it has not distanced itself from certain pitfalls that raise concerns, particularly on topics such as compliance, ethics and governance.

• bias Prejudice – Eliminating bias is necessary to more accurately reflect society. This requires identifying all potential areas of bias and calibrating AI solutions to address them. If not trained on a variety of data sets, a biased AI system will provide biased insights and recommendations.
• Loss of Control – With the increasing use of artificial intelligence, machines have become more capable of making important decisions. However, it is still necessary to involve humans in any decision-making process that may affect them in any way. AI is still unable to correctly interpret emotions, apply empathy to situations that require it, pass moral judgments, or achieve creative results.
• Technology is not foolproof – innovation is an ongoing endeavor, and if decisions are entrusted entirely to AI systems with little oversight or calibration, there is always a risk of potentially serious errors. Technology is not perfect after all.
• Privacy – Privacy has long been a major ethical issue related to artificial intelligence. For example, your smart device is constantly taking cues from its environment, such as speech, which can then be mined for insights and recommendations. AI-based toys can also collect data on children, which is a real concern.
• The erosion of trust – The unlimited collection and storage of sensitive data such as biometrics raises questions about trust – can those in charge do anything else with our private data?

Mere compliance is not enough

Addressing these issues requires us to go beyond mere legal compliance. Instead, various factors such as privacy, human rights, and social acceptability must be considered. This type of problem solving should not be limited to businesses involved in the development and marketing of artificial intelligence. Issues related to AI must be addressed across the entire supply chain – including individuals and organizations providing AI-based services.

These include:

• Compliance with the EU Artificial Intelligence Act, the U.S. Artificial Intelligence Risk Management Framework, China Internet Information Service Algorithm Recommendation Management Regulations, Singapore Artificial Intelligence Ethics and Governance Knowledge System and other global Relevant laws and regulations.
• Develop principles to guide and recognize respect for human rights as the highest priority at all stages of business operations related to the use of artificial intelligence.
• Facing social problems caused by the use of artificial intelligence in the future through the use of technology.

Technologists are not the only stakeholders when it comes to solving artificial intelligence problems. Policymakers will also play an important role in helping us address the potential risks of AI applications. They will be responsible for:

• Strengthening existing AI regulatory and governance frameworks.
• Advocate for trial and error, repeated testing, and sandboxing to fine-tune and calibrate AI models and systems to eliminate the risk of bias.
• Develop industry-specific regulations tailored to various applications of artificial intelligence in various industries.

Finally, investments must also be made to ensure that the ecosystem remains viable and that there is a deep and robust talent pool to continue staffing AI-related roles. Consulting firm Korn Ferry estimates that Asia-Pacific's TMT industry may face a talent shortage of 2 million people - including AI professionals - by 2030, with an annual opportunity cost of more than $151.6 billion.

For example, Singapore has committed $180 million to accelerate AI research and launch programs to upskill those with AI skills. While these may involve "hard" skills such as artificial intelligence engineering and development, "soft" skills are also important. Nanyang Technological University in Singapore and the Computer Society of Singapore have also successively launched artificial intelligence ethics and governance courses to recognize and certify professionals in these fields.

Great promise, but more progress needed

Ultimately, there needs to be a deep and open discussion about what AI can and cannot do. Organizations and governments alike need to ensure it will be used to benefit as many people as possible. Steps are needed to ensure that AI does not intentionally or unintentionally ignore certain subsets of society, mitigate privacy and trust issues, and implement safeguards that give humans a degree of control.

The increasing use of artificial intelligence in daily life will undoubtedly continue to raise questions about ethics, compliance and governance. Are humanity’s AI goals ambitious or simply dangerous? This is a question that all relevant stakeholders – governments, regulators, innovators, technology companies and consumers – need to work together to answer.

The above is the detailed content of The Importance of Ethics and Governance for Artificial Intelligence. For more information, please follow other related articles on the PHP Chinese website!