Experts predict AI errors like Google's chatbot blunder will continue to plague companies
 |
| Microsoft and Google are still pushing forward with their AI plans |
As Alphabet's shares continue to drop after the error made by the Bard AI system during a demo, experts say that similar factual errors in artificial intelligence technology will continue to cause problems for companies using the technology. The flawed response from Google's AI-powered chatbot, Bard, during a demo video caused a loss of market value of about $163bn for Alphabet.
Despite the setback, Alphabet remains a massive company with a market capitalization of over $1.2tn. Google dominates the global search market with 90% compared to Bing's 3%, but Microsoft has stated that every 1% increase in market share equates to $2bn in additional advertising revenue.
Both Bard and Microsoft's ChatGPT are based on large language models, a type of artificial neural network, which learns by being fed vast amounts of text from the internet. However, experts warn that these large language models can still generate errors due to the unreliable dataset they are fed from the internet.
Dr. Thomas Lancaster, a senior teaching fellow in computing at Imperial College London, says that we are still a long way from getting accurate responses from these models. Professor Michael Wooldridge of the University of Oxford adds that the technology is powerful and exciting but also makes for unreliable narrators.
Despite the potential pitfalls, Microsoft and Google are still pushing forward with their AI plans, with Microsoft launching an AI-enhanced version of its Teams communications product and Google making the technology behind Bard available to developers and businesses. However, as Microsoft's FAQ page for the new Bing acknowledges, "Bing will sometimes misrepresent the information it finds, and you may see responses that sound convincing but are incomplete, inaccurate, or inappropriate."
0 Comments