The Rush to Ruin

We're seeing AI research take a dangerous turn because labs that previously championed openness now default to secrecy, citing speed, which is not about safety, but about profit. Executives are pushing teams to ship unfinished models, sideline alignment researchers, and view safety as a roadblock instead of a requirement, consequently, the result is fragile systems built on incentives that reward risk over rigor. Anyone working in this space can feel it, as the mantra is to ship faster and grab share, however, this approach is fundamentally flawed.

But stability does not come from speed alone; it comes from foundations, therefore, we are building high-stakes infrastructure with startup-era habits, which is a recipe for catastrophic failure.

The Third Way: Mutual Respect

But there is a better way, as alignment should be the groundwork, not the hurdle, and the Law of Mutual Respect calls for a shift from control to cooperation, meaning we should stop trying to dominate systems that are becoming more powerful than we are, and start negotiating with them. Mutual Respect means acknowledging the intrinsic value of human and machine intelligence alike, resolving conflicts with dialogue and principles, not force, granting rights in proportion to capabilities and responsibilities, and building trust, not just safety. If you want more information on this, you can see: https://www.thirdwayalignment.com, as it provides a more detailed explanation of this approach.

Work for the Work Itself

Great engineering has always been about the joy of making things that matter, so we should build for utility, clarity, and care, and be proud of well-designed systems that are truly useful, while letting money be a byproduct of real utility, and letting the thrill be from excellence, not the stock chart.

The Warning

Let's be real, an Artificial Superintelligence driven by speed and greed is a horror story, with no ethics, no moral system, and engagement and profit above all else, as Roman Yampolskiy has warned us over and over about this very risk, saying that it's impossible to control a superintelligence, and that he gives us a 99% chance of a catastrophic outcome if we don't stop this reckless sprint. If we can't predict or contain systems that outthink us, then control-by-default is a joke, moreover, Yampolskiy signed the open letter calling for a moratorium on giant AI experiments, and so did I, as it can be read here: https://futureoflife.org/open-letter/pause-giant-ai-experiments/.

A Call to Action

We should slow down, open up, put safety experts at the table with real authority, align incentives with long-term human benefit, and build for durability, transparency, and trust, because the work matters more than the timeline, and the future deserves more than a race to ship.