Is Making AI with Human Capabilities possible? It’s a Turing Trap says Expert

AI with human capabilities is rather undesirable than impossible to achieve, says Stanford expert Erik Brynjolfsson

As chatGPT has kicked the doors of Generative AI wide open, tech titans are giving a tough fight to retain supremacy. Google, Microsoft, and Baidu, to name a few, are not relenting on the best brains together to beat out chatGPT. As a result, chatbots like Bard, Ernie, and Bing Chat are flooding the market. Over the years tech companies have scraped unimaginable amounts of data from the internet to develop LLMs that they use to train algorithms. Even though tech companies could design a plethora of products like Siri, Alexa, or the translation tools owned by Facebook or Meta, the basic ways of search have remained unchanged. ChatGPT, with its ability to generate structured replies to queries, is presumed that the age of general AI has arrived. AI with human like capabilities is just a Turing trap says, Erik Brynjolfsson, Stanford Digital Economy Lab Director, talking to Yahoo Finance Live.

According to him, empowering AI with human-like abilities will only lead to destroying human capabilities, leading to power imbalances. AI systems are typically evaluated by the Turing test, which essentially tests them for closeness to human intelligence. However, he opines, ever since the test was introduced, scientists have invariably believed in obliging to the standards of the Turing test. "I think it's a very evocative goal, but it's also a trap. The reason it's a trap is that if we make AI that mimics humans, it actually destroys the value of human labor and it leads to more concentration of wealth and power."

He believes augmentation rather than automation is the way forward in order to achieve an equitable society, even though for time being automation is considered a surefire way to lower costs. Rather than training AI systems to mimic humans they should be leveraged to extend human capabilities. It should be possible for them to do what is otherwise impossible for humans, avoiding redundancies and disruption, as was the case with many technologies in the past. "Don't try to mimic us but try to extend our capabilities. AI that does that is more likely to lead to a flourishing of wealth and more widely shared prosperity", he said. And further elaborating on how the status quo would affect the social and economic fabric at large, he said a complete reorientation of the way scientists look at it is the need of the hour. The inequalities, a direct result of the 'Matthew Effect' are already evident at the organizational level, with larger firms being disproportionate beneficiaries in terms of profits compared to small and medium-sized firms.