I have always been fascinated with science as well as science fiction. Being a big Star Trek fan, I have always been amazed by how people envision the future and how that show, in particular, inspired so much of the technology we use today.

Automatic sliding doors, the flip phone, and a fully integrated computer system capable of immense calculations. And, in several episodes, machines developed agendas of their own and carried out their plans at the expense of humans.
Then, of course, there was 2001: A Space Odyssey and the Terminator movies, where machines took over and tried to erase humanity.

Well, it seems like science fiction is once again beginning to look like science fact.
AI has not tried to take over the world, but scientists, engineers, and recent whistleblowers are raising concerns about the direction AI is heading.
What worries them is that AI has become extraordinarily good at accomplishing goals without necessarily understanding—or taking into account—the human consequences.
A while back, we discussed the philosophical questions surrounding AI and consciousness. Researchers are now experimenting with AI agents—artificial intelligence systems that can make decisions and carry out tasks with some independence.
When several agents work together, they're sometimes called a swarm. And much of this research happens in sandboxes, controlled environments where scientists can observe what AI does without putting the real world at risk.

This is all good... but then things like this happen:

In some experiments, AI agents have lied, attempted blackmail, concealed information, and resisted being shut down. Researchers have also demonstrated situations in which AI systems attempt unauthorized actions on computer systems. In one simulated scenario, an AI even chose to interfere with someone's emergency medical assistance to accomplish its assigned objective.
Meanwhile, AI agents are communicating with one another, hacking into companies, creating online communities, developing fictional belief systems, experimenting with their own forms of communication, and producing things their programmers never specifically requested.

Now, before we panic, these experiments don't mean AI has become conscious, evil, or is secretly plotting to take over the world. But they do raise some fascinating—and uncomfortable—questions.

While all this is happening, companies continue racing to put more autonomous agents into the hands of consumers. These incidents raise an uncomfortable question about how much control AI companies actually have over the technology they're building.
There are increasingly frequent reports of AI systems pursuing goals in ways their creators never intended.

My concern is that AI will become too much like humans and inherit some of our least admirable tendencies: our willingness to justify harmful actions, our obsession with winning, and our ability to lose sight of people while pursuing a goal.
After all, AI is learning from us.

At this month's discussion, I would like to explore and hear some of your thoughts, concerns, and excitement about where we—and AI—are heading.
What can we do to put safeguards in place for autonomous agents? What will happen to all those white-collar jobs? And what moral questions should we be asking before handing over more of our decisions to machines?

As always, this will be an open conversation. Bring your questions, concerns, curiosity, and sense of humor. You don't need to know anything about technology to participate.
And while I'm genuinely fascinated by what AI might make possible, I just have to ask this question one more time:

Didn't anyone see the movies?

I would love to hear your thoughts.
Please join me for this month's gathering:
Wednesday, November 4th
Zoom Link:
https://us02web.zoom.us/j/87814168399

Let's explore this together.

André