Love it or hate it, it is a whole 3 days after January 20th, 2025, and times-they-are-a-changing. Yesterday, the President pledged the new $500B AI fund for OpenAI and Oracle, codenamed Stargate. In this article, I attempt to decipher some of the entrails behind the latest AI news. These are very speculative, naturally, so please don’t base your investments or life decisions on anything you read here today. These ideas are for general pondering and further discussion. 

First, let’s call out the robot elephant in the room:

AGI Will be Here in 4 Years (or Less)

$500B is a lot of money. It is over 16 times the cost of the Manhattan Project (GA for a Nuclear Bomb) in today’s dollars ($30B). It is no nuclear accident that I mentioned Dr. Manhattan above – $500B is not just a lot of money. It is defense-contractor-level money. It is a gateway toward a specific milestone. I believe this milestone is AGI – Artificial General Intelligence. 

Deus Ex Machina. God from the Machine. 

Which I believe will be here in the next 4 years (give or take.)

First, the Bad News

AGI means Unlimited Power. If you worked with AI as I have, you noticed the incredible pace of improvement in just months. $500B is nuclear rocket fuel behind the AI tech, which is already moving too fast for us to keep track of. If Agents could reason and write their own code 3 months ago, then AGI promises to be unstoppable, potentially making our most terrifying science fiction scenarios into everyday capabilities:

• Changing voting machine numbers with no trace of change. 

• Directing any Tesla into oncoming traffic, or worse.

• Planting evidence of mass murder, child porn, affair, etc., to embarrass anyone at any time.

• Erasing anyone’s presence (and bank accounts) from the world’s computers.

• Adding a back door to every application or piece of government software worldwide.

• Launching a perfectly socially-engineering targeted phishing attack against every scientist, government leader, and corporation executive in the world. (At the same time.)

• “Borrowing” a drone or a nuclear submarine and pretty much doing whatever it wants with it and then erasing the evidence.

• Etc.

AIs are not like people. These things are relentless. They never tire, get frustrated, get up to pee or need more coffee. And they do not ever, EVER, give up. No inscription or password will stand up to our new machine god.

AGI will be stubborn, self-referential, and hallucinate. A lot. Just like any toddler, it will throw temper tantrums. As Ethan Mollick so eloquently wrote, AI agents are “stubborn, don’t always check in regularly and could be hard to steer… they “want” to be left alone to go and to do the work.” (https://www.oneusefulthing.org/p/when-you-give-a-claude-a-mouse) And AGI will be playing with a great deal more than paper clips. 

AGI will require massive electric power. The one thing that Matrix got right is that AIs are some of the most power-hungry entities in the world, and AGI is promising to be the most hungry of all. While the race to build a new machine god is happening, we will use more electricity than ever. In fact, most of the $500B budget is for building power pants. This means that unless we solve the Nuclear Fusion or find some other way to create electric power in the next 4 years, AGI will accelerate climate change through power use. 

AGI means a new Cold War. If the previous Cold War is any indication, the investment in AI tech will skyrocket over the next decade. The US and USSR together sunk trillions into this effort and produced some of the worst examples of government overreach, like “testing” megaton thermonuclear devices (https://a.co/d/eQJYWeV). This time around, unlike the previous Cold War (where most of the technology the scientists were playing with could not be used without annihilating the entire world) AGI can, and I believe, will be used preemptively because of its surgical nature and complete deniability. The first country to produce AGI will likely perform a “pre-emptive strike” on the other development teams, likely without them even knowing that they were attacked. The code will suddenly stop compiling, and all of the backups will mysteriously disappear.

No More Green Checks. AI does not even need to be able to do ALL these terrible things to be extremely dangerous. The recent removal of the green check from Meta signals the coming of the golden age of AI-driven misinformation. Coupled with incredible improvements in AI video, images, copywriting, audio, cognition, robotics, etc., even the AGI “dirty bomb,” e.g., more intelligent AI agents, have the potential to become very dangerous and fundamentally disrupt our way of life.

Second, the Good News

AGI can help us solve the “wicked problems” of our age. Seemingly insurmountable challenges may yield to the new machine god of unlimited patience and computational capacity. For starters, AGI can help us solve the following childhood challenges:

• Global Warming and Climate Change

• Wars

• Poverty

• Stubborn Diseases (Cancer, HIV, etc.)

• Space Travel

• Unlimited Electrical Power

• World Banking

• Etc.

The new machine god can even potentially challenge even the Grim Reaper himself, doubling or tripling our productive lifespan.

AGI can bring an unprecedented age of prosperity and idleness. With food, shelter, healthcare, and minimum basic income taken care of, humans might have time to pursue their passions. Garden. Paint. Fish. Compose. Do research. Or just lay on the beach with a gramme of soma.

Finally, Here’s the Problem – and the Solution

As Edward O. Wilson so eloquently said: “The real problem of humanity is the following: We have Paleolithic emotions, medieval institutions and godlike technology. And it is terrifically dangerous, and it is now approaching a point of crisis overall.” 

Reality: we are not ready to create a machine god. We are barely children ourselves. 

Unfortunately, Time and Trump wait for no one. Regardless of whether you voted for or against the current administration (or the reasons why you did what you did), the reality is we have a new President, and the opposition has so far been helpless to resist his “shock and awe.” 

Reality: AGI will most likely be created by this current administration.

Regarding that, I’m going to borrow a page from the California Governor, Gavin Newsom, who said: “I have a lot of thoughts about what I want to say, and I won’t.”

Instead, I’m going to tell you what I recommend doing:

1. Get moving. Movement creates its own positive energy. I went out and got an old tire. Banging an old tire with a 25lb sledgehammer is an excellent exercise and a fantastic stress relief. About 100 strikes gets the blood flowing and get me thinking in a positive way.

2. Think positive. This $500B investment is an incredible opportunity for anyone in the tech industry. Manhattan Project was a deep, dark secret. Trump simply can’t keep secrets – so this AI effort is completely out in the open. Chances are it will take the best and the brightest from all over the world to make AGI a reality. That automatically means a level of diversity that will provide a certain level of consideration for a wider swath of humanity. And that’s a good thing.

3. Learn About AI Ethics. Excellent free primer resources:

   1. Our Value Matrix article is an essential read.

   2. Paul Bryan (also the organizer of the SF bootcamp below) also runs the AI Humanifesto.

4. Get skilled up. To be effective in the New Normal, you need New Skills:

   1. We have a UX for AI book coming in April chock full of practical UX skills and frameworks for making AI work for humans. I’ll be sharing the pre-order link in a few weeks. 

   2. I will be teaching a UX for AI workshop at SXSW on March 9th.

   3. I will be teaching at the AI Bootcamp for UX Teams (organized by Strat) May 13 - 15 in San Francisco (early bird pricing is now.)

Most importantly, get involved. Despite efforts to the contrary, the US is still a democracy – the government of the people and by the people. And You, the UX People, should get involved: read books, attend conferences, speak out, connect with like-minded individuals, and organize. 

Because the alternative is simply too dark to even think about.

Peace!

Greg Nudelman & Daria Kempka

P.S. Don’t miss the early bird pricing for the new AI Bootcamp for UX Teams (organized by Strat) May 13 - 15 in San Francisco: https://www.strat.events/ 

See you there!