“Singularity” is achieved when human and machine become united. This would become possible when a computer’s processing power is integrated with human intelligence, or when computers become so complex that they could think. An exponential growth in intelligence so that centuries of progress can be achieved in months could become possible when singularity is achieved. Singularity was recognized in the 1950s. John von Newman, a computer scientist, said that the “progress of technology” would lead to “singularity.” In the 1960s Irving John Good, a British mathematician, said that “the survival of man depends on the early construction of an ultra-intelligent machine.” Ray Kurzweil, a computer scientist, said in the 1990s that “artificial intelligence is near.” In 2005 he said that “artificial intelligence is nearer.”
OpenAI is the vehicle that developed artificial intelligence. It is expected that this vehicle will establish singularity. Sundar Pichai, the CEO of Google, has described it as “more profound than fire or electricity or anything we have done in the past.” Bill Gates, the co-founder of Microsoft, suggests that artificial intelligence will change the way people “work, learn, travel, get health care and communicate with each other.” Microsoft and Google are investing billions of dollars in the further development of A.I.
ChatGPT, the artificial language model driven by AI, was created by OpenAI. ChatGPT was launched in November 2022 and by January had 100 million users. Currently ChatGPT responds to questions and can compose written content including letters, essays, articles, emails and other material. Other chatbots, like ChatGPT, are being created. The current singularity debate is being fueled by the “large language models” (LLMs). When a question is asked of an LLM it can interpret the question, determine its response, answer speedily, and with illumination.
Fears that AI will destroy humanity are being expressed. It is suggested that in the future governments or other powerful agencies can develop advanced AI systems to deal with business or warfare and when efforts are made to shut them down, they could resist. More realistically, companies could design autonomous AI systems to control infrastructure, power grids, stock markets and military weapons. While such developments would pose potential dangers, they are not quite realistic at present. What is realistic, and has been happening, is the creation of systems that can take action. The AutoGPT is such a system which is under creation. It is hoped that it would be able to generate computer programmes so as to be able to do anything online such as retrieve information, use applications, create new applications, even improve itself. Systems such as AutoGPT do not work well now but could in the future. This is the reason why realistic alarm is being raised about A.I. and calls are being made for regulation.
Not being openly discussed is the vast profits being anticipated. The promise for the immediate future is that businesses will be able to replace existing systems for faster ones. However, those who will benefit are the already wealthy corporations and businesses that can reduce cost and improve the bottom line. The information revolution has not economically benefitted working people in the wealthy, industrialised, economies. The disparity in income levels has grown exponentially since the 1970s and the level of income for working people has not increased since the 1970s. A.I. will benefit the rich minority, not the struggling majority. It was pointed out that developments in A.I. might enable retail shops to dispense with checking out staff. This will certainly enable the stores to make more profits but there will be no benefit to the customers and staff will lose their jobs. Developments in information technology over the past fifty years have certainly brought about revolutionary change, and the lives of millions in developing countries have improved. But the economic circumstances of the vast majority of people in developed countries have remained static.
Experts and observers have already recognized the enormous windfall that Silicon Valley companies will reap over the coming decades and are calling for the imposition of creative tax regimes so that the struggling majority will benefit from the developments in artificial intelligence by way of an equitable distribution of the additional trillions that will flow to the companies. The familiar arguments will no doubt ring in the ears of the needy, namely, that incentives by way of high profits must remain in order to ensure that that adequate research is undertaken and that job creators should not be penalised. In the meantime, more billionaires are created and the first trillionaire will emerge.
A.I. has given rise to creative developments in familiar areas. One lawyer told me that he fed the relevant information into ChatGPT and out came a letter of demand. It took ten minutes and he was able to charge his usual half an hour fee. One lawyer in the US submitted legal authorities in a case before the court that he obtained from A.I. The legal authorities turned out not to be in existence. Some courts in the US are considering a requirement that lawyers certify that authorities that are cited have not been obtained from A.I. (I acknowledge articles by David Streitfield, Daron Acemoglu, Simon Johnson and Cade Metz for the information and analysis contained herein).
(This column is reproduced with permission from Ralph Ramkarran’s blog, www.conversationstree.gy)
