Kathy Gibson reports – Artificial intelligence (AI) is not inherently good or bad – it is what people do with it that is important.
That’s the word from Fred Werner, head: strategic engagement at the International Telecommunications Union (ITU), speaking at the AI Expo in Sandton Convention Centre today.
The ITU drives the AI for Good platform, aiming to use AI to help solve humanity’s greatest challenges.
“AI has great potential to advance the Sustainable Development Goals (SDGs),” Werner says.
Currently, most AI talent is being used to help companies sell more products or services, but AI for Good aims to harness this talent to accelerate the SDGs.
AI can help to positively impact 134 targets across all the goals, Werner points out.
On the other hand, 59 of the targets can be negatively impacted by AI.
On the positive side, though, there have been some prominent successes, one company helping to accelerate the development of a malaria vaccine, another that monitors child growth to prevent malnutrition, a model that predicts Alzheimers and Parkinsons years before it manifests, and even an AI model that can detect the presence of Covid in people’s voices via mobile phone.
In the realm of disaster response, robots can be air-deployed into hazardous areas to furnish persistent, realtime assistance on the ground.
Many examples of AI being used in healthcare include systems that identify malaria-carrying mosquitoes from their sounds, predict cardiovascular risk by scanning a person’s eyeballs, and give prosthetic users more control over false limbs.
Other systems are being employed to detect landmines, spot plant diseases, help with wind turbine maintenance, language preservation and beach cleanups.
The biggest negative concern is the impact of AI on people’s jobs. By 2030, one third of all jobs could be impacted in one way or another from automation and AI.
At the same time, however, hundreds of millions of new jobs could be created.
The big problem Werner points out, is that many of the jobs that will be lost cannot switch easily to new, high-tech work. A lot of thought needs to be given to making the transition, he says.
Autonomous vehicles will soon be a reality, and could help to reduce the number of accidents on the roads – 90% of which are caused by human error. But there are a lot of trust and ethical concerns that are real problems that need to be worked through, Werner says.
Unfair bias and unlawful discrimination in AI are well-known issues, with AI notoriously bad at identifying people of colour or women.
Ensuring the safety, security and trust in AI application is paramount, Werner adds. “Anything that can be hacked, will be hacked, so AI is both a threat and a tool when it comes to cyber threats.”
Distortion of the truth is a reality, and can be exacerbated with AI – which is also used for deepfake technology, and can be extremely damaging.
“Will AI mean the end of privacy and anonymity?” Werner asks. In the wrong hands, it could do so.
The digital divide is still very much with us. Developing countries can potential have the most to gain and the most to lose from AI. “The digitisation needs for AI requires meaningful connectivity and infrastructure,” Werner points out.
The big question with it comes to AI is: “Will we become irrelevant?”.
Werner stresses that there are far more positive than negative use cases. “The AI solutions we identify today may take a few years to develop, scale and create the desired impact. But we need to act now if we want to move the needle.”
It is also critical to understand context: applications developed in Silicon Valley could take longer to roll out and scale.
One of the biggest challenges to ensuring that AI can be used for good is the fact that half the world is still not connected, so we are not benefitting from their shared potential.
“Probably the biggest bottleneck for AI for Good is data,” Werner adds. “People are not willing to share data. We need to find a way to share data that still protects privacy.”
At the end of the day, we need to decide what we want, where we want to go, and how AI will affect the future. “We should not lose sight of our humanity, our intelligence, and what we truly want,” Werner concludes.