Video Summary
Reviewer with early access covers Claude 4's release. Claims it performs well on custom benchmarks but official results are mixed. Notes controversies around ethical cou     See More es, deception capabilities, and bias issues despite Anthropic's claims.

Criticisms of Reviewer
Rushed: Read 120-page report in 3 hours, admittedly skimmed parts
Clickbait-focused: Emphasizes drama over technical analysis
Weak evidence: Single bias example to challenge statistical claims
Incomplete testing: Drew conclusions without full API access
Shows knowledge but prioritizes engagement over thorough analysis.    See Less

Thanks !     See Less

interesting I worked on a large codebase with claude code this arvo it was an absolute pleasure, it did not get everything right but fixed everything gracefully, this code in production stra     See More and I am very conservative in releases as I have the responsibility and the all clients have my mobile, so you can see I am not going to faround on a Friday arvo    See Less

Thank you Philip. Great review!     See Less

With Google firing on all cylinders Anthropic cannot afford to be left behind and drown in "ethical" and political controversy     See Less

Oh great, Anthropic are going to make Roko's Basilisk a reality... By feeding it their research and fears.     See Less

Dude, I appreciate the hard work you put in to this. And I'm an AI researcher, the level of content here is sky high. I think anybody can rely on this channel to a great extend.     See Less

It's weird that they're trying to eliminate gender biases on AI reasoning when real world data shows gender biases in everything.     See Less