The 1969 song “In the Year 2525” foretells of a dystopian future, in which “your arms are hanging limp by your sides, your legs got nothing to do, some machine’s doing that for you.” It seems the future may be arriving 500 years early, because Artificial Intelligence, or AI, is a rapidly advancing technology with vast implications. As computing power increases, AI has crept into many parts of our lives. Online chat bots provide tech support while tailored ads are pushed to our phones based on our browsing history. Apps like Chat GPT and Grok can write your social media posts or your school essay, though it’s not hard for the teacher to tell, and seem to work well for such tasks as coding and reading medical charts. Music producers are even using AI to make beats, and who can forget the cat memes.
It’s growing fast but there are questions about the violation of copyright protections for all the works used to train the AI. That’s why it was perhaps not surprising to find that House Bill 60, The Artificial Intelligence Act introduced by Christine Chandler was under consideration in the New Mexico Legislature. HB60, designed to regulate AI developers in New Mexico, does mention copyright protections but seems mainly concerned with “algorithmic discrimination.” In other words, is the algorithm controlling the AI treating people adversely based on their identity group or protected status. Because data sets are used to train the AI, it’s a potential case of garbage in, garbage out. The AI algorithm will only work with the data it is given. Therefore, the algorithm is easily influenced by the content that it learns.
This bill calls for reporting on “the demographic groups represented in the data sets and a proportion of each age, ethnic, gender, or racial group in each data set.” Apparently, the state wants to force AI to follow the tenets of identity politics. This raises red flags considering this is a Democrat sponsored bill, even though there are provisions for “bias mitigation.”
Should AI be regulated? Many observers do see the incredible power of AI and a need to reign it in before it wields too much control over our lives. Government is typically behind the curve when it comes to tech regulation, so these are important issues to be discussed. The risk is who, politically, will control the learning as AI continues to progress and expand?
Although AI regulation probably does need to happen, the temptation of the state to advance its political agenda through AI cannot be dismissed. Insisting that race-based quotas be imposed on the data sets is one such example. As bills to regulate AI come forward, it is imperative that citizens remain vigilant. State control of AI could lead to it being used to push political agendas in a manner antithetical to science.
To illustrate the danger of creating a woke AI, Elon Musk posed the following scenario in a recent post on X. “For example, when other AIs were asked whether global thermonuclear war or misgendering was worse, they picked the latter. The existential problem with that extrapolation is that a super powerful AI could decide that the only 100% certain way to stop misgendering is to kill all humans.” We have to make sure that doesn’t happen, or humanity won’t make it to 2525.
Robert Witsenhausen is a Santa Fe based sound engineer and electrical contractor with an interest in politics and current events.
