Our popular culture has generally typecast Artificial Intelligence (AI) as the evil villain. In the movies, the plot typically involved humans becoming overly dependent on a super intelligent AI, which realizes it is smarter than its’ human creator, becomes self-aware and decides to take control and chaos ensues. Hollywood also likes to give the AIs the launch codes to our nuclear arsenal. Remember the Terminator movies? A malevolent AI named Skynet with access to our nuclear launch codes unleashed nuclear Armageddon on the human race fearing humans were going to shut it down — the so-called Judgment Day.
AI has suddenly dominated the news cycle and we have been bombarded with a full discourse on AI’s technological benefits and AI as an existential threat. AI tech executives presented before Congress, Wall Street projected billions of dollars will be made on this new technology and the media provided over the top coverage.
What triggered all this attention? Last November, OpenAI, the public purpose corporation whose mission is to ensure AI benefits all humanity, released Chat GPT, its’ AI Chatbot text application, generating over 100 million users in only a few months making ChatGPT the fastest growing application in tech history.
I am here to report that AI Armageddon is not around the corner but a bell has been rung. I visited the OpenAI site and signed up for the free service (there is a $20 per month premium tier). ChatGPT has a Prompt window similar to a Google search window. The product uses a Large Language Model (LLM), a type of AI that’s very good at predicting what word should come next to the point where it understands context and can produce the material requested for things like topic reports, business letters, poems, etc. I created a “Chat” and entered a prompt asking for information about sci-fi movies that had AI as the villain. In seconds the information appeared.
Rather than simply provide me a list of publisher content sites to visit like a typical Google search, the ChatGPT search information appeared as a summary report. Compared to a Google search, it did a lot of heavy lifting since I did not have to visit each site to gain a basic understanding of the subject matter. This is only true as long as you trust the information being generated. I followed up with a more refined Prompt and hit the “Regenerate Response” button and more detailed information appeared. It was Google on steroids and I liked the way it interacted or chatted with the user.
In 2015, OpenAI was originally formed as an “open source” (publically accessible source code) non-profit company in contrast to a closed source, for-profit company like Google. An open source free model is a noble endeavor but hard to monetize. The co-founders committed $1 billion and included superstar tech entrepreneurs: Reid Hoffman (LinkedIn), Peter Thiel (PayPal), Sam Altman, OpenAI’s CEO (Y Combinator) and Elon Musk (Tesla, SpaceX), among others. In 2018, Musk resigned from the OpenAI board and is now a leading voice talking about the dangers of AI. In 2019, OpenAI converted into a hybrid “capped for-profit” company to be able to grow, compete and attract investors. Earlier this year, Microsoft invested $10 billion in OpenAI, a declaration of war on Google and other tech giants. Google was caught flat footed by the OpenAI media explosion, which was odd since ChatGPT took direct aim at Google’s core search business. Google quickly rushed out news about its AI product called Bard. Microsoft says OpenAI software is now being integrated into its Bing search engine (a distant #2 search engine behind Google for decades) and into other Microsoft products.
Is AI Pandora’s Box or a Magic Treasure Chest? In addition to text, AI is rapidly expanding to include video, voice, graphic and photo applications and will only get exponentially smarter and scarier. Wall Street says AI will be one of the greatest technological advancements in human history that will increase productivity, GDP and eliminate tasks but not jobs. The reality is that AI will be very disruptive and eliminated millions of jobs including white-collar jobs. Think about your own business experience and imagine all the jobs that could be done by a very, very smart AI assistant. State and Federal government will have to develop innovative public policy to cope with this coming job carnage and avoid what happened when tech innovation and foreign competition eliminated millions of blue-collar jobs and government was asleep at the wheel. Does anyone believe Congress is up to this task?
Sam Altman, OpenAI CEO, his company now center stage, appeared before Congress and basically begged for AI regulation. Now that is rare. His visit may have been contrived to buy time and create some good will. He knows there is 2024 presidential election coming and if you were appalled by what happened on social media during the last election cycle imagine what AI-powered lying will produce. If blamed for the shenanigans AI could gain super villain status.
Regulating AI will also be hard. The EU will lead as it generally does on all tech regulation but it needs to be a worldwide effort and how effective can it be without China. At the very least, all AI content should be labeled as such, include a list of its source materials and should always respect the content owner’s intellectual property rights. Unfortunately, I see decades of copyright infringement litigation. Content owners, in response to yet another attack on their business model by tech giants, will claim AI is stealing their intellectual property and demand compensation.
It is not Judgment Day yet but I read that Super Intelligent AI is coming by 2049. Best you hide the nuclear launch codes now.
Hugh Panero, a tech & media entrepreneur, was the founder & former CEO of XM Satellite Radio. He has worked with leading tech venture capital firms and was an adjunct media professor at George Washington University. He writes about tech and media for the Spy.