AI Deveopment Is Not Going To Slow and One Possible Outcome Is Manged Chaos (LIke What We Have Now)

Say what you will about President Donald Trump, the new politcal norm is that you just don’t know what to expect next and you just have to be prepared. That has been the political reality and that is how I see where we could be head with artificial intelligence. And this past weekend, we were all thrown a curve ball when Dario Amedeo, CEO of Anthropic suggests the frontier labs slow down the rate of development out of ear that AI ability is at the point where they pose a risk (he was sort of vague) to security (but he framed i as safety) and have the ability to improve upon itself, recursive self-improvement.

Recursive self-improvement is going to be something we will hear a lot about in the coming weeks as politicians, AI supporters and critics debate over the call to slow things down.

I like to dive deeper into this. It is important to point out that Sam Altman, CEO of OpenAI and Elon Musk (in all fairness has preached about this for years). For Altman, I was surprised by his willingness to join in he call to slow things down whereas. With Musk, I always assumed that he wants things slowed down so that he can try to catch up with his own AI models. Grok is behind, after all and is not considered a frontier model.  

Need to least, it is not going to happen. China is not going to comply, Meta and SpaceX is not going to be slowing down conserving that they are behind. “Sure, you (meaning Anathropic and OpenAI) should slow down. We’ll use this time to catch up.”

Here is one likely scenario that will come as a result of this. We will hear about how the new Seder development modles are progressing so fast and  companies are trying to keep guardrails on so that when they are finally released to the public, hey are safe to use and will not be misused. 

• It also means that the public will have the best models that these companies are willing to release but they will be behind anything the labs have and would have released at the previous pace. 
• China will have to be contend with stealing, I mean “distill”, from publicly released private models. Their pace of development will slow dramatically. Less and less Depseek and Kimi moments. 
• AI demand will continue. Ada center constructions will not abate.
• Dotcom like bubble talk has diminished but the attempts to talk down AI and its usefulness will continue. Just watch YouTube and you’ll see how many ad and Sbustack dollars there are.

If the threats posed by AI does come to pass as suggested by Anthropic, this is likely what will happen. There will be chaos as a result but it will be managed by the labs and companies that provide cyber security against AI attacks. Let’s not kid ourselves. Bad actors have already looked into using IA to do bad things or have already attempted it. 

Just like how we were all shocked by companies reporting how their databases were hacked and customer data were leaked or stolen, we will be bombarded by the bad things that hacker groups and countries like China, Iran, North Korea, and Russia use AI seams to attacked Western interests. And over time, we will become numbed to that as well. There will be chaos but managed. 

I think it is more important that the US and allies worth with these companies to make sure we stay as far ahead of obese bad actors as much as possible. 

For our part, as the public, we need o say engaged. I am in the camp of what is best in our interests. I don’t trust these tech bros either. They are going to keep pushing the envelope. This talks a bout slowdown is , as some have suggested, an attempt to shift blame when the coming AI chaos happen. They can freely put on the guardrails on these models if they want. No one is stopping them.

Remember that these guys have companies to run and an AI race that will never end. Whether it is he US vs China or Claude vs Gemini or ChatGPT, no one can afford to lose. Dario calling for this at this moment is suspect. But it does remind us that AI is here to stay - there will be benefits and there will be times when things go bad. How badly and how we deal with the consequences is entirely up to us.