(RightWing.org) – When it comes to technology and computers, what is meant by the singularity? Actually, it’s been discussed quite frequently as of late, as in artificial intelligence (AI) and the continuing advances in programs like ChatGPT and IBM’s Watson. One definition of it is the point “where computer programs become so advanced that AI transcends [outstrips, beats] human intelligence.” Another of the baby steps towards that potential has recently been announced.
Moving Along
One of the rapid up-and-comers in the field, OpenAI, recently announced that they are “rolling out voice and image” features for their ChatGPT program. They say this will allow users to have a discussion with their program rather than it simply replying with a text response to their inquiries.
The announcement also indicates they will soon be adding a feature that will allow the application to recognize input via photographs. They even provided three hypothetical uses, which are:
- Photographing a landmark and having a live back-and-forth with the bot.
- Uploading pictures of the contents of one’s refrigerator and pantry and being provided with a recipe.
- Providing an image of a child’s math homework with a particular problem circled, with the AI then sharing hints to solving it.
The Washington Post interviewed one of OpenAI’s vice presidents, Peter Deng, who told them that the vocalizations will be available in five new personas, with each having “a different tone and accent.” The article notes that it is this type of feature that has some experts warning that real-life people may begin to anthropomorphize the machines, thereby assigning too much trust in what they are hearing.
The believability of the voices that are generated could pose a real problem because all the bots on the market are given to what is termed “hallucinations.” Since they are now programmed to respond to open-ended questions, they try to predict what the user will say next, leading them to invent false narratives.
Putting that in perspective for today’s America and the upcoming 2024 presidential election cycle should put a chill of fear down any person’s spine. When one considers the relatively clumsy automated social media posts attributed to foreign countries in the 2016 and 2020 elections, it’s not unreasonable to wonder what catastrophic damage could be caused by a completely conversational human voice spreading fake news reports.
Congress has taken up the idea of regulating AI with agreement from both ends of the political spectrum, such as it is and as far as it goes. Republicans and Democrats alike have publicly said it is a necessity in the future in order to protect the integrity of elections, ensure that civil liberties/rights are not infringed, and stay ahead of the People’s Republic of China (PRC) and their global ambitions. However, beyond that, it seems that the typical partisan split between Liberal overregulation and Conservative open market ideals remains in full force when it comes to the details.
Copyright 2023, RightWing.org
