Sakana AI has progressed beyond its early demonstrations. But I did not find evidence that it has established a widening research lead through sustained, accelerating RSI. Its published results support bounded self-improvement and increasingly capable orchestration. Your test asks for something stronger: does the improved researcher make the next improvement faster, cheaper, and more consequential?
So far the faster and faster on RSI methodology is drowned out by faster and faster making and raising more money and bigger and bigger AI Data centers and hiring the best staff. A company can build faster and faster like say SPACEXAI with data centers and is more efficient on the build. When does algorithmic RSI loops take over. I have a combat move that gets 1.2X better and more powerful everytime I use it but then I get knocked out because someone bought the current heavweight champ. The supposed critical and most important factor is not even overwhelming most of the other factors. Not even 10% weighting of factors.
Useful RSI does not require accelerating returns. A system that repeatedly improves itself while gains gradually diminish can still create substantial value.
A widening competitive lead is neither necessary nor sufficient. However, the stronger claims that we can apply AI to go faster adn faster means at some point a leader in this area would be able to compound gains to pull away. Competitors could improve equally quickly or copy published methods. Conversely, a company could pull ahead because it bought more compute or hired stronger researchers.
Stronger self-accelerating research—the evidence should look like this
Freeze model access, total budget and evaluation standards.
Compare each researcher generation on fresh, consequential problems.
Measure accepted improvement per dollar and elapsed week.
Attribute gains to inherited changes, separating external model upgrades and human redesign.
Show repeated gains after promotion into the next researcher.
Confirm those gains survive integration into real systems.
Sakana has demonstrated pieces of a self-improving research team. It has not yet publicly established that the team is becoming better at producing its next breakthrough at an accelerating rate.
Brian Wang is a Futurist Thought Leader and a popular Science blogger with 1 million readers per month. His blog Nextbigfuture.com is ranked #1 Science News Blog. It covers many disruptive technology and trends including Space, Robotics, Artificial Intelligence, Medicine, Anti-aging Biotechnology, and Nanotechnology.
Known for identifying cutting edge technologies, he is currently a Co-Founder of a startup and fundraiser for high potential early-stage companies. He is the Head of Research for Allocations for deep technology investments and an Angel Investor at Space Angels.
A frequent speaker at corporations, he has been a TEDx speaker, a Singularity University speaker and guest at numerous interviews for radio and podcasts. He is open to public speaking and advising engagements.
