There is an elephant in the room.

Actually, judging by this week’s television interviews, headlines, and op-eds, the elephant has escaped the room, booked a publicity tour, and is appearing on every Sunday morning program available.

The elephant, of course, is artificial intelligence—and whether it is going to kill us all within the next ten years.

Former Anthropic researcher Jacob Coxon recently warned that the people building advanced AI genuinely believe it could destroy humanity by the end of the decade. Anthropic CEO Dario Amodei—whose name already sounds like someone who should be delivering an ominous warning from a candlelit castle—has been making the rounds talking about the dangers of rapidly advancing AI. OpenAI CEO Sam Altman has also joined the conversation about slowing development and putting stronger safety measures in place.

Apparently, Terminator 12: The Chatbot Has Had Enough is already in preproduction.

Let me say this before someone accuses me of welcoming our new robot overlords: I am not suggesting that artificial intelligence comes without risks.

Of course it needs safeguards. Of course there should be oversight. Of course the companies building increasingly powerful technology should not be allowed to regulate themselves entirely behind closed doors while the rest of us simply trust that they know what they are doing.

History has given us more than enough evidence that “Don’t worry, the corporation is monitoring itself” is rarely the beginning of a comforting story.

But there is a very large space between taking AI seriously and announcing that it may exterminate the human race before some of us finish paying off our furniture.

That is where I begin giving the entire conversation my Sunday Side Eye.

I use AI regularly. I use it to brainstorm, research, organize my thoughts, refine ideas, create content, solve problems, and occasionally confirm that I am not the only person who finds something completely ridiculous.

AI is a tool in my business. It isn’t my business brain, my personality, my experience, or my judgment. It doesn’t replace the 26 years I’ve spent learning how people communicate, market their businesses, make decisions, and occasionally make complete fools of themselves on social media.

It helps me work. It doesn’t run my life.

That everyday reality seems to be missing from much of the public conversation. We are repeatedly offered two choices: either AI will create a glorious future and solve every problem known to humanity, or it will become self-aware, take control of the internet, launch the missiles, and leave one lonely Roomba circling the remains of civilization.

Could we possibly discuss the enormous middle ground?

There are real problems happening right now that deserve attention. AI can make scams more convincing. It can help spread misinformation faster. It can imitate people’s voices and images. It may eliminate or significantly change certain jobs. It raises serious questions about privacy, copyright, bias, surveillance, and what happens when a small number of extraordinarily powerful companies control technology that millions of people depend upon.

Those aren’t imaginary science-fiction scenarios. Those are current issues affecting actual people.

But “We need thoughtful laws, independent safety testing, transparency, and accountability” doesn’t attract quite as much attention as “THE MACHINES MAY KILL EVERYONE.”

The more dramatic version gets the headline, the television segment, the interview clip, and the social media argument. Fear travels well. Nuance usually misses its connecting flight.

And perhaps the AI leaders giving these warnings sincerely believe every word. I’m not inside their meetings. My magnifying glass has its limits.

But these warnings also create a strange contradiction. The same industry developing the technology at breakneck speed is now standing before the public saying, “You have no idea how dangerous the thing we’re building might become.”

Well, gentlemen, you’re the ones building it.

If the risk is truly that immediate and catastrophic, perhaps another op-ed and a Sunday morning interview are not quite enough. Stop racing. Agree on meaningful safeguards. Invite independent experts inside. Support enforceable regulations. Explain the risks in language ordinary people can understand without requiring us to choose between blind trust and an underground bunker.

Don’t terrify the public on Sunday and return to the development race on Monday.

I believe we can hold two thoughts at the same time: Artificial intelligence is powerful technology that deserves careful oversight, and the claim that it will kill us all within ten years is not an established scientific fact.

It is a prediction based on what future systems might become, what capabilities they might develop, and what humans might allow them to control. Possibility matters, especially when the potential consequences are enormous—but possibility is not certainty.

Aliens might also arrive within the next ten years. Given the size of the universe, I’m not ruling them out. But until one of them lands on my front lawn and asks for the Wi-Fi password, I’m going to keep the situation in perspective.

So yes, let’s talk about AI.

Let’s demand transparency, safeguards, accountability, and a healthy amount of human judgment. Let’s pay attention to the real risks without either worshipping the technology or turning every conversation into the opening scene of a disaster movie.

And could everybody please get a grip?

The elephant is already taking up enough space without giving it a flamethrower and glowing red eyes.

Sunday Side Eye
Looking at business, technology, and social media from behind the magnifying glass—because somebody has to.