King Charles has warned of the “existential dangers” posed by AI.
I suppose that makes a change from the time he warned that the “doomsday clock of climate change” was ticking toward midnight. Or from when he told us that genetically modified food risked “the biggest environmental disaster of all time.”
I do not like to criticize the King of my native country, but I am not sure it is wise for him to keep straying into public policy like this.
To be fair, the King is not the only one warning of the dangers of AI. Similar warnings have come from inside the industry itself, from leading figures such as Dario Amodei and Sam Altman.
Why, though, would people who have invested billions of dollars in cutting-edge AI want to invite government regulation of innovation in their own industry?
Over the past few years, Western AI firms have poured vast sums of capital into building foundational AI technology from scratch. Alarmingly, there is evidence that a number of Chinese AI firms have been able to replicate much of what those companies built. The FBI has described this as “industrial-scale” copying of American intellectual property.
Having failed to protect their advantage by technical means, the industry now wants the state to protect it for them.
Remember how, a couple of decades ago, we all used Yahoo before Google came along?
Imagine if Yahoo had been able to persuade the US government to restrict the development of search engines. What if Yahoo had warned of the risks of “irresponsible” search engines that would allow anyone to look up anything? Some people, they might have said, could even look up how to commit dastardly acts. Regulate it!
Or remember when everyone had a BlackBerry, rather than an iPhone? What if BlackBerry had persuaded the US government to slow down the development of smartphones? Those touch-screen devices, they might have warned, would allow all sorts of unsavory social media. Best control it!
When AI industry leaders urge the state to regulate their sector, I do not believe they are being altruistic. They are lobbying for a regulatory “moat” to protect their commercial position. If they really wanted to slow down, they could slow down.
Among the commentariat, I cannot help noticing that some of the most alarmist voices on AI are the same people who told us Covid was going to kill us all.
Cast your mind back to those days of hysteria. Millions of people in countries such as Britain were placed under effective house arrest, needing a reason even to leave their own homes. Transatlantic travel was suspended for months. Schools were shut. Businesses were closed.
The real disease was not the virus. It was a form of neurosis: a hysterical overreaction that did far more lasting damage than the virus itself.
In the 1970s, the doomers told us that nuclear winter would end humanity. In the 1980s it was acid rain. In the 1990s it was global warming, which during the noughties morphed into climate change. Then came Covid hysteria. Now it is AI.
I have no doubt that there will be downsides to AI. Like the printing press, new technology can be put to good purposes and bad ones. The printing press allowed the Bible to be printed and read in many languages. The same technology allowed the publication of Mein Kampf.
AI, too, will serve both good and evil. Human ingenuity will require further ingenuity to counter the evil. The risks of AI will need to be tackled — by AI. Just as bad ideas in print were countered by good ideas in print.
I am sad to see the King adopt, yet again, such a doomster stance. Think of the history of the English-speaking peoples at times when there really was reason to lack confidence in the future.
Think of Charles’s forebear, King Alfred, in 878 AD, when Christian England seemed on the brink of extinction. Or in 1315, when there really was an environmental disaster: the crops failed and famine followed. Or imagine what it must have felt like to live in England in 1348, when the Black Death killed four in ten people.
You want doomster? Imagine what they felt in November 1916, when the Battle of the Somme ended with some 420,000 British casualties.
You want pessimism about the future? Imagine it is London in October 1940. Your city is being bombed to rubble and a Nazi army is massing for invasion.
Except that instead of doom, a 14-year-old girl — Princess Elizabeth, the future Queen and Charles’s mother — gave a radio address to the nation’s children. She assured them that “in the end all will be well; for God will care for us and give us victory and peace.”
Contrast that quiet assertion of faith in the future with King Charles’s endless undermining of it.
It seems to me that the further we are from real danger, and the more materially assured we become, the more we produce leaders who indulge the ridiculous idea that we are on the brink.
In the six years that I have lived in America, I have marveled at the optimism I see all around me, especially here in the South. Yes, everyone faces challenges. The future could go awry. Nothing in this world is certain. Fear and uncertainty are part of the human condition.
But here there is an inner faith in the future: that “in the end all will be well.” That is perhaps why more Americans, especially in the South, marry younger and have more children than Europeans — the ultimate vote of confidence in the future.
The West needs leaders who give us confidence in the future. Leaders who convey the idea that, no matter how great the challenge, we will see it through. Leaders who love our culture, and love those who love it too.
Yes, AI will require us to adapt — just as we once had to work out how to make trains and cars safer. But we need to stop this relentless doom-mongering. Enough of this Western neurosis.
Douglas Carswell is the President of the Mississippi Center for Public Policy and author of this post.
This post is a paid advertisement.
13 comments:
Chas was never the sharpest marble in the bag.
AI armageddon is every bit as real and dangerous as Climate Change, Net Neutrality, and the Plandemic need for the Clot Shot!
So, Douglas, and everyone else who thinks they can provide factual answers, here are some questions. If you cannot answer each and all factually and with a sentence or no more than a short-ish paragraph, you do not know what you're talking about and should probably just not say anything at all on the topic of "data centers."
1. What do "data centers" actually do?
2. How do they provide a benefit to:
a) the owner,
b) the surrounding community,
c) the country in which they are located, and
d) society in general.
3. How many data centers does the US require to function?
4. How money would it take to fund all the data centers currently operating, in any phase of construction, or currently announced?
5, How money would it take to make those in #4., above, just break even?
6. How much power would it take to operate those in #4, above?
7. What is the delta between the power required to operate those in #4, above, and the power available by Jan 1, 2028?
8. What is the range for the physical lifespan of the things that data centers must have to function?
9. What is the range of effective, "real world" lifespan of the item(s) in #8, above?
10. Where will the replacements for #8, above, come from when replacement(s) are required?
7:52 correction: 4 & 5 should begin with "How much money..."
Oh GEEZ! Doug is neither good at math or U.S. history and he certainly doesn't know crap about epidemics.
General George Washington REQUIRED his troops to have a SMALL POX " vaccination". I use quotation marks because it required exposure to a live virus.
How does Doug think we stopped polio? I actually saw my friends from behind a glass in iron lungs. We damn near stopped measles ( 6000 deaths in 1912, 3-4 million infections and 48000 hospitalizations 400-500 deaths annually pre-1963 . I have male classmates who were left sterile from teenage mumps.
We damn near eradicated both, not just from the U.S like we did measles in 2000, but stopped epidemics all over the world. My grandmother lost a son at 3 years old when one of her boys had measles and her daughter got mumps and the baby got both. Then there are the kids in my large high school who still had disfiguring pox marks on their face at our 50th+ reunions from the chicken pox. Doug, you need a decade of U.S. History classes to get you to the high school level. What exactly are your " credentials" other than 5 years and a basic citizenship test which most my kids could pass by the 4th grade. What were your majors in college and grad school? Other than you, how many members are in the MCPP and what are the dues and requirements for membership?
Nuclear power can handle individual data processing centers and energy grid issues economically and efficiently. MS, determined by some to be the armpit or asshole of the country, is a cheap site choice for AI generation and Nuke power, with plenty of startup water to loop for re-use. Strategic negotiators, like the pie forum wizards, can eliminate ingrained poverty in Jackson and MS with a made-in-heaven marriage a trois of AI, Nuke Energy and Water. Excepting Foote and Hartley, Jackson has no economic negotiators of merit, hence the necessity of bigger pie forum where energy, economics and engineering expertise thrive.
10:44AM wrote, "Nuclear power can handle individual data processing centers and energy grid issues economically and efficiently."
Explain. Show numbers and at least enough of you work to generally follow along. I cannot speak for anyone else, but I'll assume you can do it until you fail to do so.
7:52am demands answers?
Just ask AI. The better the questions are formulated, the more helpful the answers. Then go back to clutching your pearls/family jewels.
Post of the day!
Earth to 10:30. These new vaccinations are nothing like the polio cure. These are lightly tested and they change our RNA.
I asked AI, which said handling 10 large AI data centers (5-10 gigawatts) is efficiently accomplished, but adding regional grid requires a cluster of 4-6 Nuclear plants. Nuke plants can furnish the power 24/7 except for reloading every 2 years, so stagger the cluster loading among the cluster. AI tech developers are buying Nuke plants, I'm suggesting invitation for them to buy into the Nuke plants in advance and shape their engineering and location(s). Regulation and approvals might take years but Nuke power is the most efficient, cheapest, safest and continuous source of power. With advance planning and municipal/local approvals, a Nuke plant (3-4 yrs?) + AI data centers (1.5 yrs) alliance could initiate a (3-4?) years long process with short term natural gas generation until Nuke plants approved and built. Meanwhile, supporting water river dams and reservoirs can be strategically located and built. The engineers of both Nuclear Power and of AI know what works for them. The importance is to benefit the local community, the region and the grid. These parameters must be identified and negotiated equitably several years in advance by qualified negotiators who do not exist in Jackson Governance, though Foote and Hartley are common sense business minded leaders and should be included on the negotiating team, which should not include the bureaucrat/socialist minded others in Jackson governance.
When AI influences the news media, we will pay dearly.
RMQ
"Krusatyr" (Krusatyr...oh, lawdy, lawdy...) wrote:
"7:52am demands answers?
Just ask AI."
and of course,
"I asked AI, which said..."
Uh-huh. I didn't demand anything and I don't need to ask "AI" to know you don't have a clue about what you are asking "AI." Even if "AI" could figure out what you asked it, you wouldn't know whether the answers were accurate, hallucination, or plain ol' bullshit. Plus, "AI" being able to generate answers to the questions asked is immaterial to any person's understanding of the questions or the answers. Which is one of the myriad problems: little or no intelligence handling the bulk of the training for "artificial intelligence." What could possibly go wrong?
In any case, anyone who cannot provide answers to the questions asked (Krusatyr did not, even with "AI") should not be discussing data centers or AI.
Post a Comment