The AI takeover is already here, you’ve just not been paying attention – inews

author
8 minutes, 27 seconds Read
Ads
Best Mobile Games Directory

Best Mobile Games Marketplace

Artificial intelligence has been with us in some form for nearly 70 years – since the term was first invented for an academic conference in 1956. But in the last 18 months, since the arrival of ChatGPT, the chatbot developed by OpenAI, the pace of change thanks to the technology has quickened.

And that will have meaningful, long-lasting effects on our lives – most notably in how we work.

“AI is rapidly impacting all industries, and we must, first and foremost, be supportive towards those directly affected by its adoption,” said Barak Eilam, CEO of Nice, which develops AI-powered customer experience software.

The scale of workplace disruption could be huge. Investment bank Goldman Sachs believes 300 million jobs could be replaced by AI by 2030, and could boost global GDP by 7 per cent in the same time period. The International Monetary Fund sees 40 per cent of jobs worldwide being “affected” in some way by AI, with white collar roles more likely to be affected than blue collar ones.

Likewise, a new report published last week by the Institute for Public Policy Research (IPPR), a left-leaning think-tank, believes that 11 per cent of the tasks carried out by British workers could be replaced by AI – with women 40 per cent more likely to be in roles that could be replaced than men.

Glance at the headlines and AI could change our lives significantly for the worse, leaving us on the dole queue. But it’s not necessarily all bad. For one thing, AI could help unlock new jobs, alter others to make them more efficient, freeing up time to tackle more significant issues that computers can’t yet do.

“While many roles are at risk, there is also tremendous potential for job creation,” said Eilam. “Despite the urgency felt by businesses to not miss out on the AI rush, time is on our side to upskill employees across every sector in the UK – and globally – allowing them to work alongside this growing technology, rather than be replaced by it.”

Not everyone is convinced that the AI boom won’t be a workplace bust, however. While some hope that AI will help augment individual workers, enabling them to work more efficiently, others fear it will have a negative effect. “Pushing skill levels down in more and more job classifications dilutes the power of workers to negotiate with employers for decent pay and working conditions as more of their jobs become interchangeable and therefore easily replaceable,” said Mike Katell, an ethics fellow at the Alan Turing Institute, a research institute for AI named after one of the people most closely involved in its invention.

AI is expected to benefit healthcare (Photo: PA)

Beyond the workplace

One of the key benefits that many see AI having is beyond the workplace, and in sectors like healthcare and government.

Artificial intelligence, the thinking goes, can help process and analyse data quicker than humans can, getting rid of dreaded backlogs and improving the quality of services.

Medicine is one area that AI will undoubtedly change our lives. Drug companies have been able to develop new medicines at pace thanks to the ability for the current crop of generative AI tools to rattle through a vast number of potential combinations of medicines and test them in silico (on a computer). From there, the most successful likely candidates can enter human trials faster – meaning they get to patients quicker.

One British company, Exscientia, uses AI to cut the average time taken to put a medicine into clinical trials from four-and-a-half years to between 12 and 15 months.

And AI is helping on frontline treatment too, augmenting doctors’ abilities: an NHS trial of an AI tool called Mia helped identify signs of breast cancer in women that were so small, doctors had missed them when analysing mammograms.

However, how we utilise these tools will be important. The NHS trial has helped doctors’ performance because it was used in addition to, rather than instead of, their expertise at analysing mammograms. But a separate study, published earlier this month in Nature, suggested that experts could be hindered as much as helped by AI tools when analysing X-rays and CT scans.

“We find that different radiologists, indeed, react differently to AI assistance — some are helped while others are hurt by it,” said Pranav Rajpurkar, assistant professor of biomedical informatics in the Blavatnik Institute at Harvard Medical School, and one of the co-authors of the study.

Ensuring efficiency

Beyond medicine, there are plenty of ways that we’ll see AI touch our lives – in terms of interactions with government, police and the law.

“I think we will see AI more involved in the distribution of things of value,” said Jamie Susskind, a barrister and author of The Digital Republic, a book about taking back control of technology. “Increasingly our access to credit, insurance, housing, social security, even jobs, is likely to be mediated by machine learning systems.”

Your credit card or insurance provider already uses forms of AI to analyse data points about you to try and establish how risky you are as a potential customer, but the rise of generative AI, which supercharges the technology, makes it possible for them to draw on more granular data. Similarly, your bank will likely outsource its customer service to AI chatbots in the coming months and years.

Susskind points out that there are risks equal to the potential rewards from using AI in areas such as those. “It all depends on the engineering and execution,” he said. “The risks of bias and error are well known.”

Yet Susskind is optimistic about how AI could change our lives. “I would say that AI generated outcomes should not be treated as a poor substitute for human equivalents,” he said. “The legal advice may in fact be better.”

It should also be cheaper, which Susskind believes could be a gamechanger. “What AI could do, if properly regulated, is democratise access to professional services, culture and more in a way that is hugely beneficial,” he said.

But avoiding bias

Regulation will be key, because automated AI systems have plenty of issues.

Generative AI systems work by coming up with the most likely answer to questions – essentially acting as guessing tools, or predictors. It’s why a large language model (LLM)-powered chatbot like ChatGPT is more likely to say “The dog barked” than “The dog composed a Shakespearean sonnet”.

The system bases its guesses on training data it has been shown by the companies and individuals that develop it. But that training data is often only a small, biased fraction of society as a whole, rather than a representative one.

If you want to converse with an AI chatbot in any language other than English, it can be tricky to do so. That’s because the majority of the text on the internet, from which AI systems are trained, are in the English language. In all, around 54 per cent of the web is written in English, while English is only spoken by 17 per cent of the offline global population. Hindi is spoken by 3.5 per cent of the world offline, but makes up just 0.1 per cent of the web’s lingua franca.

That means for many, AI chatbots will be exclusionary. And when AI technology is used for things like policing and managing our borders through facial recognition, other issues ensue. Facial recognition systems, which are trialled and deployed by a number of organisations worldwide, can be more than 99 per cent accurate at recognising white male faces accurately – because they’ve been trained on datasets containing plenty of them. But accuracy can drop as low as 35 per cent when presented with a Black female face.

Some worry those issues and biases could become more deeply encoded by the widespread use of AI algorithms. “While there may be some improvements in advancing science and technology, the biggest risks from AI are its effects in deepening hierarchies and social divisions,” said Katell.

A surveillance camera as French police start to test artificial intelligence-assisted video surveillance of crowds in the run-up to the Olympics in Paris this July (Photo: Reuters)

In day-to-day lives

But beyond changing our working lives, potentially improving our health and being layered into our interactions with government, AI will also have other, more mundane but no less significant impacts. “The main thing that comes to mind is conversational AI, being able to talk to just about any appliance and electronic tool,” said Sasha Luccioni at Hugging Face, an AI company.

She foresees a future where the kind of chatbots that we’re seeing when using ChatGPT are integrated into home appliances, and can talk back to us. It could involve “anything from asking your oven if your cookies are ready to texting your fridge from the store to ask if you still have any milk left,” she said.

However, Luccioni also points out that wondering how AI will change your life is the wrong question to ask in many ways. It already has. “People tend to underestimate how much AI has already impacted our lives,” she said. “Toll booths that scan your licence plate automatically, navigation apps that help us avoid traffic, weather prediction… it all has AI components under the hood.”

It’s a point that Katell makes too – every time you log on to a social media platform or load up Netflix or another streaming service, you’re interacting with AI. “Apps and social media platforms are using AI-supported profiling and content generation to push cherry-picked, personalised content in the service of advertising business models and attention economies over providing broad or diverse perspectives,” he said.

That has its plus points – directing you more quickly to TV shows and films you like, while keeping those you don’t out of your recommendations – but it also results in drawbacks. “AI provides new and evolving capabilities for narrowing and personalising all manner of content and experiences and threatens shared understandings of the world during a time of intersecting global crises,” said Katell.

But for those worrying about the impending rise of AI and its impact on our living, Luccioni has a simple message: “Sorry to burst your bubble, but AI already changed your life.”

This post was originally published on this site

Similar Posts