The AI debate has developed a strange pattern.
An AI model produces unexpected or potentially scary results, someone predicts the end of humanity, and the conversation jumps straight to whether civilization has only a few years left. Missing somewhere between the demonstration and the obituary is the evidence. We should demand it.
Claims that AI will destroy humanity require a credible explanation of how software could acquire the power to overcome human resistance in the physical world. An alarming prediction, even from an accomplished researcher, doesn’t settle that question.
Longtime technology analysts should lead this discussion. Too often, we leave it to executives with products to sell and politicians who turn uncertainty into campaign material. People who understand how technology reaches the market need to speak up, together, regardless of their political views.
An Extinction Forecast Needs More Than Imagination
Start with an honest boundary. Nobody can prove that every future AI system will be harmless. But fear-mongering about an impending AI-driven extinction event treats speculation as knowledge. The stronger argument is that the evidence doesn’t justify the certainty of the apocalypse narrative.
The International AI Safety Report, published in February and led by AI expert Yoshua Bengio, found that current systems lacked the capabilities needed for the loss-of-control scenarios it examined. The report also warned that relevant abilities were improving. That supports continued testing and safeguards. It doesn’t establish an approaching extinction event.
Even a highly capable model needs a means to act. Producing a plan is different from carrying it out. Software tools can expand its reach, but access to them must come from somewhere. Intelligence doesn’t automatically grant control.
Physical constraints matter too. Computing requires equipment and electricity. Actions outside a computer depend on connections to machines or people. A serious catastrophe argument must explain how those dependencies are acquired and sustained.
These barriers aren’t guarantees. Security can fail, and people can grant dangerous permissions. But each required failure belongs in the argument. Asking for that chain of events is basic analysis.
Industry Leaders Push Back on AI Alarmism
Nvidia CEO Jensen Huang has pushed back directly. In a July interview, he criticized people who “spend too much time theorizing about these science fiction outcomes.” His broader argument was that frightening people away from AI could undermine adoption and American competitiveness.
Central to this is Huang’s message that frontier companies such as OpenAI and Anthropic should address risks through their development, engineering, and testing processes — like any product company that ships only high-quality solutions.
The standard should be clear: establish rigorous AI testing procedures and don’t release products with known critical bugs. Risk ultimately isn’t just with models. It also lies with companies that allow weak internal due diligence across their teams.
Cohere CEO Aidan Gomez has voiced similar objections. Asked in a 2024 interview about public misconceptions, he identified “the fear that certain individuals and organizations espouse about this technology being a terminator, an existential risk.”
The disagreement predates today’s models. In 2017, Mark Zuckerberg criticized doomsday rhetoric, saying “It’s really negative, and in some ways, I actually think it is pretty irresponsible.” That historical comment shows how long this argument has persisted. It cannot, by itself, settle questions about newer systems.
These executives certainly have commercial interests. Their confidence deserves scrutiny, as do warnings from competing executives. However, their objections challenge the suggestion that everyone who understands AI agrees catastrophe is coming.
Arguments for broadly shared AI benefits also offer an alternative to fatalism. Analysts should test those promises as closely as extinction predictions. Neither becomes evidence because a prominent executive sounds certain.
The Media Must Ask Better Technical Questions
A frightening prediction is easy to package. Explaining its assumptions takes work. When coverage lacks technical depth, a possibility can start to sound like a forecast. Readers may conclude the engineering questions are settled when an article has barely examined them.
Reporters don’t need to build models, but they do need to assess whistleblowers’ credibility and potential agendas more carefully. They also need to distinguish controlled-test results from evidence that a system can operate independently in the real world. Editors should demand that distinction before allowing a dramatic claim to carry a story.
Mainstream coverage isn’t uniformly alarmist. Some commentary has challenged extinction forecasts and emphasized product accountability. Other reporting has examined how doomsday speculation can distract from present harms and the people responsible for them.
The problem is coverage that repeats extraordinary claims without examining the steps required to make them possible. Technical fluency should help a newsroom recognize when a compelling story has outrun its evidence.
Data Centers Have Become the Next Target
Anxiety now extends to the infrastructure behind AI. Data centers face growing public opposition, with local construction disputes feeding a broader political backlash. Buildings full of computers are being asked to answer for everything people dislike about the technology industry.
The economic benefits of data centers are already showing up in public accounts. In Loudoun County, Virginia, existing facilities generated $1.2 billion in property tax revenue in fiscal 2026, helping fund local services and ease the burden on other taxpayers.
Across Virginia, an independent legislative study estimated that the industry supported roughly 74,000 direct and indirect jobs and $5.5 billion in annual labor income, with much of that benefit coming from construction. Those paychecks reach electricians and other skilled workers, well beyond Silicon Valley. Contracts can also create electricity savings when large customers pay more than it costs to serve them.
Georgia Power projects approximately $950 million in annual customer savings beginning in 2029 from its large-load growth portfolio, equivalent to about $180 annually for a typical residential customer. Those are projected savings, not reductions already delivered.
Every proposal needs scrutiny, but a debate that counts electricity demand while ignoring wages and tax revenue leaves out a substantial part of the economic picture.
Some objections are practical. Residents deserve answers about electricity bills and water use. Noise matters to neighbors. Developers weaken their case when they dismiss these concerns or substitute vague promises for enforceable commitments.
Still, treating every data center as a step toward human extinction turns planning into a referendum on an imagined apocalypse. Communities need to examine a facility’s actual costs and benefits.
Construction creates work for skilled trades and business for suppliers, while operating facilities can provide technical jobs and local tax revenue. Permanent staffing is smaller, and advocates should say so.
Construction jobs shouldn’t be discounted simply because they involve building infrastructure rather than writing software. Communities should examine tax concessions carefully, but rejecting a project also means giving up potential benefits.
America Cannot Regulate Its Way to Leadership
AI requires computing capacity. Excessive regulation that delays that capacity is likely to weaken America’s competitive position. Blanket restrictions and uncertain approval processes discourage investment without necessarily addressing the risks behind public concern.
Bernie Sanders’ proposal to ban artificial superintelligence and pause advanced AI development until federal safety rules are established risks putting American innovation on Washington’s timetable.
If competing countries continue developing those capabilities, an American “pause” could weaken the U.S. technological lead and create national security disadvantages. Sanders’ separate proposed moratorium on new AI data centers could also delay construction jobs and the local tax revenue those facilities can generate.
Prolonged uncertainty can discourage investment and favor large companies with the resources to wait out restrictions, making it harder for smaller competitors to enter the market. Protecting the public requires enforceable safeguards, but broad prohibitions risk sacrificing economic opportunity without securing comparable restraint from America’s rivals.
National security implications also deserve attention. The federal government explicitly connects domestic AI infrastructure with strategic capabilities. The policy implication is straightforward: making responsible construction unnecessarily difficult could undermine the capabilities Washington wants to protect.
America should remember its nuclear experience. Reactor expansion slowed beginning in the 1970s. Regulatory changes and public opposition contributed, alongside rising costs and weaker electricity demand. Blaming regulation alone misreads the history. Ignoring its role is equally misleading.
The United States shouldn’t repeat that experience with AI infrastructure. Rules should address demonstrated risks and provide a workable route to approval. Endless delay carries its own costs.
Yet AI has become a political football in a midterm election year. Complex questions are entering campaigns built around simple messages. Supporting development becomes evidence of recklessness. Supporting safeguards becomes evidence of hostility to progress. Neither accusation evaluates a model or a construction proposal.
Analysts Need to Speak Together
This is where longtime analysts must act. We are often privy to advanced product roadmaps under nondisclosure agreements, and in my experience, thoughtful companies seek feedback on their plans. We don’t have to agree on politics, but we should agree that technical claims must survive technical questioning.
A shared effort could give editors access to specialists and help them publish clear responses when claims outrun the facts. Explain what was tested. Identify what remains unknown. Ask how extinction probabilities were produced before those numbers become campaign slogans.
Authority brings obligations. Analysts should disclose vendor relationships and correct their mistakes. Experience earns attention; it doesn’t excuse weak reasoning. The independence that challenges alarmism must also challenge industry promises.
Nor should we mock worried readers. They hear warnings from respected researchers and assume those warnings mean something. Our responsibility is to explain the disagreement.
Analysts need to defend sensible safeguards while confronting exaggerated claims, across political lines and without waiting for vendor permission. We need to speak collectively and with authority on this topic, showing the evidence behind our conclusions.
Fear shouldn’t drive decisions that America might spend decades trying to reverse.
SmartTech Research analyst Brett Faulk contributed to this article.
Read the full article here
