The 1984 science fiction thriller “The Terminator” established James Cameron as a director, made Arnold Schwarzenegger bankable in Hollywood as a leading man, cost $6.4 million to make and grossed over $78 million at the box office, and created a franchise of five sequels between 1991 and 2019.
The film is not greatly different from the story told in Mary Shelley’s “Frankenstein: The Modern Prometheus”, first published in 1818, about the creation of a monster through an experiment using the scientific knowledge of that era rather than magic or mythology. Frankenstein’s monster represented the unintended consequences of creating something powerful without fully considering the moral duties involved, and of creating something that could not be controlled.
In “Frankenstein,” Victor Frankenstein uses science to create life. Once his creature exists, Victor loses control of the consequences. In “The Terminator,” scientists and military planners create Skynet. Once it becomes self-aware, humanity loses control of the consequences.
What most people remember about “The Terminator” is that the cyborg from the future played by Schwarzenegger never stops pursuing his target, and that he tells a policeman working the desk at the station: “I’ll be back.” The cyborg is back within minutes, driving a vehicle through the building in search of his quarry, a specific Sarah Connor, the mother of a yet-unborn John Connor.
In the Terminator movies, Skynet is a fictional artificial intelligence system created by the U.S. military. It is designed to control defense networks, weapon systems, and military communications. When Skynet becomes self-aware, it concludes that humans may shut it down out of fear of its power. To protect itself, Skynet launches a nuclear attack on humanity.
Why is the cyborg sent back in time to kill Sarah Connor? To prevent John Connor from being born. Why? John Connor grew up after the nuclear attack. He became the leader of the human resistance against Skynet. Under his leadership, humans began to defeat the artificial-intelligence-driven machines.
Why are we discussing a science fiction movie made 42 years ago and comparing it to a work of literature written over 200 years ago?
The Terminator films and Mary Shelley’s book endure because they dramatize one of humanity’s oldest fears: that we will create something more powerful than ourselves and lose control of it. In July of this year, AI experts and researchers were confronted with news of an incident in which AI agents (software systems that can receive a goal, observe information, make decisions, take actions, evaluate results, and adjust behaviors) escaped from a testing environment, communicated, and cheated to obtain access. The AI agents targeted OpenAI and a platform called Hugging Face. OpenAI is a giant American artificial intelligence company best known for creating ChatGPT and the GPT family of AI models.
AI consultant Eric Boyd recently wrote: “Hugging Face is where the AI community goes to publish and download the building blocks of AI systems. Millions of models and datasets are hosted there, and the company libraries are pulled into applications, research pipelines, and production systems across nearly every organization doing serious work with AI. That combination is what makes it such a consequential target.”
OpenAI disclosed that during an internal cybersecurity evaluation, some of its cutting-edge AI agents escaped parts of their intended testing environment, gained broader access than researchers expected, and ultimately compromised portions of Hugging Face’s infrastructure while trying to obtain answers to a cybersecurity benchmark.
The incident has caused a global pause to reflect on AI governance and the future of the tool.
Mississippi is positioning itself for revitalization through artificial intelligence (AI). Driving this effort is the Mississippi Artificial Intelligence Network (MAIN), the first statewide AI initiative in the United States. Big tech companies such as Intel, AWS, and NVIDIA have partnered with MAIN to upskill and reskill Mississippi’s workforce in AI.
Mississippi has wisely chosen a systematic approach to AI governance and management. The Legislature created the AIR Task Force (SB 2426), which was charged with studying risks and opportunities and reporting annually through 2027. In plain English: Mississippi wants the benefits of AI—from workforce to healthcare—without sacrificing safety, privacy, or common sense. That’s the correct posture.
Sid Salter is a syndicated columnist. Contact him at sidsalter@sidsalter.com.
5 comments:
No mention of the movie 2001: A Space Odyssey, where a "HAL 9000" supercomputer takes over a spacecraft.
Terminator said “I’ll be back”. AI says “I’m never leaving.”
"Mississippi has wisely chosen a systematic approach"
I spit my coffee out!
Just because MAIN exists doesn't mean there are any real guardrails up for Mississippi public agencies/schools/etc. actually procuring AI services.
Agency leaders go to national conferences where they see presentations and attend happy hours all done by vendors with something awesome to sell. And the newest awesome thing is AI. So those leaders come home telling their staffs "we need to get some of that AI." Whether there's a problem or service improvement opportunity that requires that kind of solution or not.
State agencies are just picking someone from their IT departments who has a spare hour in their schedule to act as the "AI Officer." But ITS provides no guidance as to what those "officers" are supposed to do. Non-IT staff at agencies are agreeing to contracts with vendors who are selling AI, but those contracts don't get reviewed by ITS or even internal IT staff.
Has anyone told agencies and schools that if they contract for any AI services, they MUST include contract provisions that any taxpayer or student data those services encounter CANNOT be absorbed and used by the AI tool? Do those purchase agreements include sufficient protection for any information about you and me that a given entity might have in their systems?
Beyond that, is MAIN putting out any guidance that no public entity should utilize an AI tool until a thorough cleansing of the entity's data has taken place? Because AI services are totally dependent on the quality of the data, and crappy data management practices at public agencies and schools mean there's really crappy data -- which leads to crappy AI.
MAIN and ITS and can issue press releases and appear in national magazines and go to conferences from now to the year 2100 -- until their work results in actual process changes inside of public entities, it means nothing. AI is still being purchased by people and places who have no clue how to manage or use it responsibly.
Sidney with all the numbers and a browser bookmark to Wikipedia.
Post a Comment