I code firmware for a heavily regulated medical device (where mistakes mean life and death), and I try to have AI write unit tests for me all the time, and I would say I spend about 3 days correcting and polishing what the AI gives me in 30 minutes. The first pass the AI gives me, likely saves a day of work, but you would have to be crazy to trust it blindly. I guarantee it is not giving you what you think it is or what you need. And writing the tests is when I usually find and fix issues in the code. If AI is writing tests that all pass without updating the code then it's likely falsely telling you the code is perfect when it isn't.

The real reason is Google has no incentive to make it smaller. In fact there is incentive to make it bigger. If your phone runs out of space you need to get a new phone with more memory, or move pictures and the like to the cloud, both of which are good news for Google. So why spend any effort keeping the size down.

Do data centers on Earth have no employees present, and none who ever come on site for the life of the data center? Prove that out on earth and I will start to believe your space data center.

I'm quite sure that can be done, if you jack up the price and pare down requirements enough. The question is, would the result be useful.

Honestly this is the best response. If the AI was actually so great, it could create better AI, and the future would already be here

They are talking to each other already.

Whether for betterment remains to be seen.

2 Opus 4 talking directly to each other: https://www.iflscience.com/the-spiritual-bliss-attractor-som...

"Learning" passing between derivatives of the same base models: https://alignment.anthropic.com/2025/subliminal-learning/

According to the Hitchhiker's Guide to the Galaxy this would take an AI 10 million years.

Seems like we're stuck with '42' for a while ;)

That's an interesting bar. What is real time? One day they are likely to be faster than us at any response.

But that is stupid. If they incentive innovation and productivity gains they could do more business. They don't have to give you the full value of your improvements to incentivise you.

Ultrasonic cutting board. "Are you tired of spending $1800 on a whole set of ultrasonic knives only to have them break every time your Brother in law throws them in a bowl of soapy water after a dinner party? Well know you don't have to. Image buying 6 cheap knives at a garage sale and turning them into chefs quality knives instantly, without even touching a sharpener!"

The real takeaway from your story is that it's easy to stay on task when you're interested in the task. Your coworker just didn't care about his work. But if his work was creating a productivity tool then he'd probably love his work and be productive.

Leap seconds are when it really gets fun.

I will never not be impressed with how well we are able to keep track of time in the world

There are plenty of non tipping ways to reward hard workers. Managers can track performance and reward it through bonuses and/or wage increases.

ive seen that work at smaller size companies (sub 2000 employees) but with a caveat - it causes some political stress for the recipient. jealous teammates are less likely to work with you after you get a bonus.

and FAANG size companies dont have random bonuses for frontlines; stack ranking is supposed to reward top performers, but when manager kpi's conflict with work that benefits the customer experience, managers will intentionally rank the best performers as average.