Dr Raymond Ashton wonders whether anticipated improvements in productivity will be as good as predicted.
The possible uses of AI have fascinated computer users for many years. The intellectual development of AI uses has exploded in recent years, and in particular in the fields of law and other media users. This facilitated the commercial development of AI and chatbot. Given the evolutionary nature of the product the possibilities of this have been discussed at some length in the popular press and in elsewhere. In this sense all sorts of possibilities have been canvassed.
Recently it was reported that a large law firm is proposing to introduce an AI chat-bot to assist them, and in particular to assist its lawyers to draft contracts and similar tasks. The aim of this much-vaunt-ed technology is to find efficiencies for both its lawyers and clients. The tool is now available to any lawyer of the firm.
While other groups have been experimenting with similar technology, other companies across other industries are exploring using the technology. This has been spurred by the November launch of ChatGPT, an AI chatbot from Microsoft-backed Open AI. This can parse text and write answers to questions.
The technology (known as large language models) is poised to be introduced to a wide range of industries that generate large amounts of text, such as media, advertising and education.
This new wave of AI systems has reawakened natural concerns about the technology’s threat to millions of jobs although in the legal field it is said that it would not reduce billable hours or save money for the company and/or its clients in the short term. The firm concerned said future versions of the technology could lead to cost reductions ‘eventually’. This move comes amid growing pressure on law firms to reduce costs, following years of rising salaries for their most junior employees. Law firms have often used technology to deliver work more flexibly but not necessarily at lower costs. The AI product comes with a disclaimer that its use should be supervised by licensed legal professionals, yet it can still ‘hallucinate’, which is when the programme can produce inaccurate or misleading results. Lawyers and others will be alerted that they need to fact-check any information generated by the chatbot. Experts have also raised regulatory concerns over the ethics of using the technology in legal settings, where accuracy is paramount.
Obviously the development of AI and chatbot methodology is in its relative infancy but it offers significant possibilities to increase efficiency in data-intensive industries. While frequent use may be 10 years away, we can expect significant improvements in the technology, the effect of which will be to greatly enhance its use. In this regard, a comparison might be drawn but it offers significant possibilities to increase efficiency in data-intensive industries. While frequent use may be 10 years away, we can expect significant improvements in the technology, the effect of which will be to greatly enhance its use. In this regard a comparison might be drawn with the internet. Indeed, since the turn of the century there have been significant improvements in productivity. One note of caution is echoed by the AI ‘godfather’, Geoffrey Hinton, who has been in the news warning that the implications of Al chat-bots are ‘quite scary’.