We know we are in the midst of an artificial intelligence revolution, which, alongside the sustainability revolution, is contributing to a period of significant upheaval. Until recently, though, the potential for the technology to transform our daily lives felt like it lay somewhere in the future.
The rapid rise of ChatGPT has changed that. The chatbot, made publicly available by owner OpenAI, is estimated to have reached 100mn users in just two months – the fastest growth for a consumer application in history.
By comparison, it took TikTok nine months and Instagram two and half years to reach the same level, according to data from Sensor Tower.
Little wonder that people in professions from journalism to medicine, computer programming to financial advice, are wondering if AI is going to replace them.
Why has the uptake been so fast, and why does it seem to herald a future that is closer than we thought? It’s because, for the first time, the technology feels… good.
With fairly limited training, ChatGPT has much more of a human feel than previous examples; from virtual assistants such as Alexa and Siri to the customer service chatbots that often seem only to compound the customer’s frustration.
Trying it myself, I asked various questions of the kind a client might ask an adviser. In its responses, ChatGPT steered clear of trying to give personal advice, but provided a wide-ranging summary of the things someone in the position I set out should think about. For a technology that, even with 100mn users, is in its infancy, it was impressive.
Unlike previous AI technologies I have encountered, it showed its workings and explained its limitations: not 'you should invest in a pension – this one' but 'if the following things are true, you should consider investing in a pension, taking into consideration X, Y and Z'.
This felt like a step forward from the black-box decisioning that has driven some of the ethics concerns about AI – the embedded bias in AI technology used in the US justice system, for example.
It is already clear that AI technology, as it is trained in specific industry contexts, will have very wide applicability. Just in the last week or so, I had a call with someone in San Francisco who used an AI to listen to our conversation and produce a summary.
I read about magic circle law firm Allen & Overy introducing an AI chatbot to help its lawyers with tasks such as contract analysis and due diligence. I heard from a top marketing organisation in the US that its new AI-generated newsletters are vastly outperforming the old ones, which were written by humans.
And on the news I saw the physicist Stephen Thaler, who is heading to the UK Supreme Court to seek a ruling that his AI can be listed as the patent holder for inventions it creates.
