By David Robb ——Bio and Archives--September 30, 2026
Artificial Intelligence, or AI, has been in the news a lot recently. Claims that it will wipe out the human race, that data centers will consume all our water and electricity, massive job losses, increased surveillance, immense loss of freedom, and many other scare stories abound. Likewise are stories of vast new opportunities, freedom from drudgery, major advances in medicine, thousands of new products, and a wide range of other benefits. Who, if anyone, is right?
Before I throw my hat in the ring, let me state that I have no financial interest in AI, I don't own stock in any AI company, nor do I work for or am I paid by any party or firm involved with AI. I did early work in AI development many years ago, and currently use AI in my own work and business. That said, let's look at some of the claims, good and bad to see what sense we can make.
First off, no-one, except perhaps a few mad super villains, want a SkyNet, or HAL, or a Terminator running loose. Objections to AI generally fall into three categories: Data centers will take all our water and electricity and privacy, AI will take all our jobs and destroy our economy, and finally, "we're all gonna’ die".
Taking the objections in order, there is a legitimate concern regarding data centers—not because of power or water use, but because of privacy and data use issues. You see, data centers serve two major purposes. The first is simply processing power needed to support AI operations. It takes around 4,000 to 5,000 square feet of server racks to support a million simultaneous AI users. Additional area is needed for processors that analyze the data to train AI models. All those servers are power hungry, and run hot so they need water for cooling just like a car.
We can expect AI use to grow quickly, so there will be a need for large processing complexes of several acres each. There are lots of data centers proposed because each AI company wants its own so I doesn't depend on competitors.
But processing power isn't the real danger. The other use of data centers is for data collection and storage. Governments, credit card companies, marketing firms, and many others collect massive amounts of data on each of us, ranging from names, ages, incomes, shopping preferences and other mundane items to things like online activity, who we associate with, where we go each day, medical information, what activities we engage in, where we bank, and many other bits of information that can be used both for and against us.
Much of this information is collected without our knowledge or explicit permission. China has institutionalized it to calculate the Social Credit scores that it uses to reward, punish and control its citizens. Many would like the same power here.
We have laws regarding collection and storage of personal information, but there are two problems. They are often not enforced, and even when they are, the penalties are trivial for large users. Many government users even excuse themselves from compliance, claiming they need the information in order to govern us effectively. The temptations are huge, and the penalties for misuse are minor—a bad combination.
Some place the blame on data centers and work to restrict them, or even eliminate them. As we shall see, while acting with the best of intentions, those seeking to restrict or eliminate data centers are misguided at best, and acting against the good of the country at worst. If I were an enemy of the US, I would do everything I could to discourage construction of AI data centers.
What is really needed are good laws with teeth to restrict data collection and retention. Permission to collect data must be explicit, data should only be retained under very limited conditions, and only for limited, specific times and purposes, and purged when those purposes have been accomplished. Penalties for misuse should include both fines that could bankrupt even the largest firms, to criminal penalties for executives and government officials alike. Once people see a few high level folk perp walked off to jail, we will see data protection and privacy laws taken seriously.
As for the second objection, yes, AI will eliminate jobs and replace human workers in certain areas. One prime area is in customer service. In this industry, outsourcing has already eliminated many jobs, sending the work to countries where English is barely a second language, and product knowledge is no better than that of the person seeking help. Replacing these jobs with AI has to be considered a net gain.
Other, more skilled jobs have been considered as candidates for replacement as well. Some firms have tried replacing software developers, for instance, with AI programming systems, often to their detriment. Good software development requires human understanding of not only the direct problem but also of the environment of a program's use. The actual coding of a program is a relatively small part of the effort, and one where AI in the hands of a skilled developer can be put to good use. Many firms that mistakenly thought they could get rid of their software engineers are now seeking to re employ those they let go.
AI is a powerful tool with the potential for profound change to society, much like steam engines in the Industrial Revolution, or the automobile more recently. Both of those examples caused substantial changes in the nature of work and society, but also brought about vast new opportunities.
Fears that AI will eliminate some jobs are very real, but most jobs will survive. Many will change and many new ones will be created—jobs we can't even imagine today, just like computer technician jobs were unimaginable a hundred years ago.
Consider the automobile. Just a few years after its introduction, millions of horses became unemployed, along with thousands of people who cared for them, fed them, and cleaned up after them. Saddle and bridle makers, carriage makers, and livery stables shut down permanently. On the other hand, thousands of new part manufacturers started, millions of mechanics, thousands of service stations were built, road construction boomed, auto dealers flourished, and a myriad of other new jobs were created, most of which had never even been imagined in the time of the horse.
Social changes were of similar magnitude. Where people lived, what jobs were available, even dating and mating practices changed. People traveled to places in a few minutes or hours that would have taken days or weeks before.
All this illustrates one of the major problems of any significant new innovation. The problems and difficulties are quickly obvious and often painful, while the benefits and opportunities take time to develop. Many of the opportunities will be discovered or invented over a period of years, so even if they result in benefits far greater than the original disruption, they often go unrealized or taken for granted.
AI is in the same position today as the steam engine or the automobile, or even airplanes were soon after their introduction—the pain and problems are evident, but the benefits are just beginning to show. What is needed is a good way to deal with worker displacement, not a way to cripple or prevent introduction of one of the most powerful technologies of the age.
As for the third objection—we’re all gonna’ die—there is reason for concern, but hot because of AI. Aside from the fact that everyone dies eventually, the human race has always been faced with existential threats of flood, famine, war, pestilence, cosmic disaster, asteroid strikes, to name just a few. Not so long ago, we were worried about nuclear winter.
Today we have alarmists claiming the AI sky is going to fall. They offer no credible explanation why AI systems would want to wipe out humanity outside of perhaps our potential nuisance value. When questioned in greater depth, most of these people are worried about two things: making people redundant and unnecessary, and people using AI to create things that could wipe out humanity. Powerful bioweapons are often examples used.
As for making humans redundant, Jack Williamson, a science fiction writer, wrote extensively of the dangers of intelligent robots. His novel ‘The Humanoids’, along with his earlier novella 'With Folded Hands' from 1947 addressed many of the same issues we hear today regarding AI. Not to dismiss concerns, but humans and AI systems have far more to gain from a form of partnership than from either wiping out the other.
As for misuse of AI to create WMDs, AI makes the perfect watchman to identify and defeat such efforts. The question "Quis custodiet ipsos custodes?”—who will guard the guardians?—goes back over two thousand years. Today AI can watch for the sort of developments, including the use of other AI systems in development work of things that could have mass adverse consequences and raise awareness in time to address them.
That is not to say that some extremely clever super villain might not, in James Bond fashion, try to create some such super weapon, or some science fiction mad scientist with a beautiful daughter might not invent some world-destroying invention, but I think we have more to worry about from an asteroid strike. Just ask the dinosaurs.
View Comments
David Robb is a practicing scientist and CTO of a small firm developing new security technologies for detection of drugs and other contraband. Dave has published extensively in TheBlueStateConservative, and occasionally in American Thinker.