@MJTruthUltra Gee I wonder what you would get if you place that background image next to its mirror image , flip the whole thing 180 degrees and play with the contrast ...
👉 Continuous Re-evaluation Over Rigid Planning
Jensen rejects formal planning cycles: No 5-year or even 1-year plans.
The idea is constantly reevaluating strategies based on changing business and market conditions.
This is very useful given the rapid pace of AI development.
👉 Open Access to Information and Decision-Making
Everyone in the company should have full context, all the time. No meetings are exclusive to VPs or directors.
Jensen believes in sharing strategic directions with everyone: "Why just tell one person?"
Jensen doesn't use status updates because he believes they are too refined when they get to him.
Instead, anyone in the company can email him their "Top five things" regarding whatever is at the top of their mind, and he will read it.
He reads 100s of these every morning.
It can be useful, if you can afford to run each experiment multiple times.
On the other hand, statistical tests are shockingly often wrongly used. ANOVA should not be used when even slightly deviates from Gaussian, since it'll lead to really misleading p-values. I was quite shocked to learn how, in medical research, people would spend years of manpower and millions in funding, only to throw out the results based on a wrongly chosen significant test. Use non-parametric tests, people.
@svpino OpenAI probably has to serve a lot more requests. 4o feels like it was watered down to handle the scale. I expect Claude to push a degraded model at scale too. At some point non degraded model will be a premium subscription.