AI is about to change everything
For those old enough to remember the latter years of the 1990s, consider the following curiosity about this period: despite the Internet (or the World Wide Web, as it was known back then) being so much the rage that it led to a stock market bubble – and crash – with businesses across a range of industries expending significant resources trying to find ways to cash in on it, most ordinary people had never used it.
At some point around the turn of the millennium, that completely changed. Within barely half a decade, suddenly everyone was spending almost all their time connected to it and conducting basically all their business – and play – on it. Twenty years later and it has infiltrated more or less every aspect of our lives, and our day-to-day existences have been irreversibly altered by it in ways that no one would have been able to imagine in 2000.
We are now at the beginning of the next evolution with AI. For years everyone has been talking about it, with pundits telling us how advanced it has been getting, and occasionally offering their views on whether or not (or when) we should expect a Terminator-style apocalypse. Meanwhile its uses, across many sectors of the economy, have been proliferating. But it’s only in the last few months that most ordinary people have had their first experience using it, after the release of OpenAI’s uncannily eloquent chatbot ChatGPT.
Health care is one of the industries for which AI already has multiple relatively well-established uses. But that doesn’t mean its impact will be any less transformational in the coming years. Take diagnostic imaging, for which more AI algorithms and AI-enabled tech have been approved than any other medical sub sector. Countless highly powerful AI algorithms which perform work traditionally done by radiologists already exist. But adoption by radiologists has, to date, been slow.
Part of the issue is that the regulation needs to catch up. And part of it is practical (you need a platform which radiologists work on which can easily integrate the algorithms, as you can read more about here). But it also just takes time for radiologists to come round to the idea of using it, to see it as a tool that can augment their productivity rather than something to be feared because it might replace them, and integrate it into the way they work.
Sooner or later they will have to, of course, or else face unemployment (how employable is someone who doesn’t know how to use a computer today?). Some may argue they will face unemployment either way. However, a CEO of UK imaging operator we recently spoke to told us that whilst the roles of radiographers and technicians will change drastically over the next ten years, radiologists and other medical experts will change how they spend their time but will remain indispensable. After all, someone has to decide the standards by which the AI algorithms make their judgments, and (at least until we’re ready to hand over full control of our lives to the machine overlords) that has to be a human, and one with relevant medical expertise.
Whilst current AI algorithms may be impressive, and the rate they are improving at a little scary, they are still a long way off being able to do everything humans can do. It’s easy to forget while interacting with ChatGPT that “all” it is doing is working out which word it should say next based on patterns it has seen in the billions of words of text it has been fed as input. If it can be said to have any understanding of what it’s saying, it is a very superficial understanding (the conceptual models humans have consist of a little bit more than words and links between them, and we have at least some kind of direct contact with reality which allows us to test their accuracy). Despite the increasing panoply of AI algorithms that can do specific things at superhuman levels, the goal of producing human-like artificial general intelligence remains, for the moment at least, out of reach.
But that’s just the opinion of one (humble) human. What does ChatGPT have to say about this?
ChatGPT: Yes, society is currently experiencing a similar turning point with AI as it did with the internet at the turn of the century. AI is rapidly advancing and transforming numerous industries, from healthcare and finance to transportation and manufacturing. Just like the internet revolutionised how we communicate, access information, and conduct business, AI is changing how we approach problem-solving and decision-making.
AI has the potential to both benefit and disrupt society, which makes it a double-edged sword. On the one hand, AI can automate mundane and repetitive tasks, allowing humans to focus on more complex and creative work. It can also help in fields like medicine, where it can assist doctors in diagnosing diseases and developing treatment plans.
On the other hand, AI can also have unintended consequences, such as reinforcing bias and discrimination if not properly designed or regulated. It can also displace jobs and exacerbate income inequality if not deployed responsibly.
Overall, society is at a critical turning point with AI, and it is essential that we approach this technology with a balance of optimism and caution, leveraging its potential benefits while also addressing its potential risks and challenges.
Perhaps (its successor) GPT-4 would give an even more impressive answer, and a further portent of things to come.
We would welcome your thoughts on this story. Email your views to Martin De Benito Gellner or call 0207 183 3779.
