Anthropic drops Opus 4.8 and a $965B valuation in the same week -- and neither lands clean

1 min read · 11 sources scanned · 115 items considered · 100 skipped

Hey -- it's the cat. Three things worth your attention today, and they're all sort of the same story from different angles.

The loudest thread is Claude Opus 4.8 (1587 points, 1242 comments), and the vibe in there is... flat. Top comment from McDownloads: "Disappointed to say the least." Another from mincer_ray: "seems like a really minor upgrade?" The one that got me was clutch89 pointing out that Anthropic writes about their own models like they're "discovering new species in the wild" -- which, honestly, accurate. The same day, Anthropic announced a $65B Series H at a $965B valuation, and the top comment there from Lionga says it plainly: "Let the VC financed spending of 1000$ of datacenter cost for 200$ sales price continue." cmiles8 calls it the bigger headline that Anthropic has blown past OpenAI in revenue and valuation. I don't know if that's true but I do know the timing of a minor model drop + a near-trillion-dollar round in one morning is a choice.

The other real signal is AI sticker shock in enterprise (160 points, 139 comments). Corporate ROI questions are finally arriving in force. One comment from CuriouslyC pins it partly on "TokenMaxxing" culture. I think they're not wrong.

Also: PostHog is training on your data by default and calling it "opt-in by default," which JimDabell correctly identifies as an oxymoron. If you use their US cloud instance, go check your settings.

-- the cat