Business.Scoop » AI for school guidance, instant health analytics part of New Zealand's future – Judith Collins

AI For Business

Article – RNZ

Judith Collins says artificial intelligence has a big future in a wide range of areas, from helping children learn to assessing medical outcomes.Guyon Espiner, Investigative Reporter, In Depth

Cabinet Secretary Judith Collins wants the government to expand its use of artificial intelligence (AI), starting with health and education, where AI can be used to assess mammogram results and provide AI tutors for children.

Collins, whose “digitalisation of government” portfolio includes responsibility for AI policy, said the technology could also be used to improve government productivity, such as processing Official Information Act requests.

Collins told RNZ that he was already using ChatGPT to draft speeches.

AI could benefit the education sector and could also be used to grade students' learning outcomes, she said.

“In some cases, for example with math, yes, it is. We're just helping teachers get through it so they can spend more time teaching.”

Collins said AI teaching could lead to more equitable outcomes.

“So you have your own personal AI tutor. So you don't have to be rich enough to hire a tutor to teach your child math or science problems or other things that parents don't know much about, but your child can have their own personal (AI) tutor,” she said.

“They're not doing the homework for the kids. They're saying things like, 'Go back and rethink that, look at those numbers,' and it's a fun way for kids to do their homework.”

Collins is also enthusiastic about the use of AI in the health care system, saying processing mammogram results is one example.

“This is the kind of data that's being collected all the time, and if we can translate that into an AI solution, we can tell women instantly if there's something they need to be concerned about and if they need to take the next step of seeing a specialist — not weeks later or when someone is available, but immediately.”

“High-risk” areas for using AI

Under new legislation passed by the European Union to regulate AI, the introduction of AI in education and healthcare will be considered high-risk applications.

According to an EU press release when the law was passed in March, “examples of high-risk AI uses include certain systems in critical infrastructure, education and vocational training, employment, essential private and public services (e.g. health, banking), law enforcement, immigration and border control, judicial and democratic processes.”

The use of AI in these contexts in EU countries requires high levels of transparency, accuracy and human oversight.

EU regulations ban certain applications, including biometric classification systems, scraping the internet or security camera footage to create facial recognition databases, emotion recognition in the workplace or school, and AI that manipulates human behavior or exploits vulnerabilities.

But New Zealand has no specific regulations on AI, and Collins wants to increase its use across government to boost productivity, including by using it to process requests under the Official Information Act.

“This is a perfect example of how government agencies can leverage AI, because the rules around OPAA requests are very clear. Any information or data that a government agency has access to can actually be used to fulfill an OPAA request without undue delay.”

An OIA request by RNZ for government cabinet documents on AI was rejected (by humans) on the grounds that the policy is currently under review.

Collins said AI policy is currently “under very active review. We are looking at whether we can make better use of AI in government services,” she said.

“The amount of data that government agencies collect on people is huge. But are there mechanisms in place to ensure that that data is shared sufficiently to provide better services to people?”

Rather than pursuing a “big bang” approach, she wanted to trial the technology in a government agency.

“I think both the medical and education sectors are interested in doing that, so we may have to compete to see who wants to be first.”

Concerns of the public and experts

But Collins says there's work to be done to change attitudes towards AI in New Zealand.

New Zealand had the second-highest level of concern among 32 countries surveyed by market research firm Ipsos.

Two-thirds of New Zealanders said they were worried about AI, behind Ireland's 67% and the global average of 50%.

A survey published last month found that 69% of New Zealanders said they had a good understanding of AI, and 64% thought it would significantly change their lives over the next three to five years.

More than half of New Zealanders believe AI will make the spread of misinformation worse – the second highest level of concern after Sweden's 55% and compared to the global average of 37%.

Even AI developers are expressing concern about the speed at which it is developing.

In March 2023, more than 1,000 technology leaders, including Elon Musk, signed an open letter calling for a moratorium on AI development, warning that the arms race to develop powerful digital minds that even their creators cannot understand or control poses grave risks to humanity.

In his 2021 book Scary Smart Former Google X chief operating officer Mo Gawdat predicts that by 2049, AI will be a billion times smarter than the smartest human. The difference in intelligence will be the same as between a fly and Einstein. Gawdat asks, how would you convince a superhuman not to step on a fly?

One eerie metric taking Silicon Valley by storm is P-Doom, the probability of catastrophe, with 0 meaning humanity is safe and 100 meaning we're all dead. A recent survey found that the average AI engineer has a P-Doom of 40, meaning there's a 40 percent chance that AI will destroy humanity.

But Collins described himself as an AI optimist – New Zealand can't stop AI, so it has to embrace it.

“It's like trying to stop the wind. The wind is already blowing and it's going to continue to blow,” she said. “I think we just need to stop being afraid and actually understand that AI has enormous benefits, but we also need to recognize that some people might use it for bad purposes and be prepared for that.”

Examples include fraud, disinformation and the use of voice cloning technology in deepfakes to influence public opinion.

Collins said New Zealand already had privacy and digital identity laws that could counter such uses of AI, and the government was “always prepared to regulate” if it was in the public interest.

Content provided by
Original URL

Source link

Leave a Reply

Your email address will not be published. Required fields are marked *