OpenAI CEO and ChatGPT creator Sam Altman believes that despite the obvious benefits of AI tools is not yet dangerous, but "I think we are not that far away from potentially scary ones."
The Dangers Of Unregulated AI
One of the dangers of AI Altman mentioned is a bias for or against particular political beliefs. While much of the bias could simply be due to data, experts have warned that there could be some inherent algorithmic bias against marginalized groups, based on AI developers mostly being from relatively similar socio-economic and racial backgrounds.
Recent examples with ChatGPT have also shown that the technology can often furnish incorrect facts with a degree of confidence that can lead some people to believe in their accuracy. This is especially troubling at a time when fake news and misinformation are at an all-time high. When Altman spoke about the 'scary' aspects of the technology, he likely meant some of the issues that have already cropped up over the past few years, even though the technology is still at an early stage. That includes privacy violations due to social surveillance with AI, as well as the increased instances of deepfakes that have wreaked havoc. It could potentially become almost impossible to distinguish between credible information and propaganda.
Apart from rampant misinformation, other challenges because of the emerging new technology are large-scale job losses and a new-age arms race of AI-powered weaponry. While the ubiquitous use of AI seems almost inevitable now, developers and regulators have to ensure that they plug as many loopholes as possible, enabling people to enjoy the benefits of the technology, without suffering from its unintended pitfalls.
Source: Sam Altman/Twitter