Sir Nodebridge
1.9K posts
@chessieowl
Chess, startups, tea ☕, and a bit too much Agatha Christie
So many people do not understand this - and it is core to consumer software👇 Gen Z and beyond does not care about data privacy as long as a product delivers for them Bank accounts, contact book, email access, camera roll, etc. - they will gladly give it up
Anthropic's SMB launch told us exactly where they're not going to compete... yet. They built horizontal connectors into HubSpot, Stripe, QuickBooks, Gmail. This is useful for the 10% of SMBs who run their business on horizontal tools. The other 90% don't live in HubSpot. They live in vertical software. Or pen and paper. And Anthropic just showed their hand. There were two strategic options. 1/ Sell to every vertical platform and take margin on inference. 2/ Or go direct, build the wedge, own the operator. They picked neither. Which means the operator relationship in every vertical is still up for grabs. The second wave of vertical SaaS will win by letting the operator build the software themselves.
I see a lot of problems with this and it is kind of scary. But, I would buy it. This would be super useful.
We’re all figuring it out
all I want is for my notes/content repo to be: 1. simple, reliable, fast 2. take recordings and do transcription/ASR 3. legible to code agents 4. enable more content creation (.md guidance) w/repo 5. not trap me in weird proprietary agent-building gui 6. sharing/permissions
You've been asking for this one... Now in preview: Codex in the ChatGPT mobile app. Start new work, review outputs, steer execution, and approve next steps, all from the ChatGPT mobile app. Codex will keep running on your laptop, Mac mini, or devbox.
Stop turning prompting into magic spells (and yes, this includes random slash commands with obscure outcomes). Let this one area of working with AI not be weird. Just ask for stuff, in well-specified formats, like a manager, not a sorcerer with a bunch of incantations.
Most TTS benchmarks come from the vendors themselves. Conditions get picked to flatter the system being measured, using studio-quality text, simple inputs, and P50 only without spread The numbers are real. They're just not comparable. And they don't tell you what happens when 5% of your calls hit a 600ms tail. We built benchmarks.coval.ai to fix that. Standardized conditions, open methodology, open code, continuously updated. Today's snapshot: Gradium ranks first on both latency metrics we measure (P50 TTFA and P25-P75 spread) without trading off intelligibility. Their WER is competitive with the top of the field. Full per-provider breakdown is in our benchmarks page. Run it yourself: github.com/coval-ai/bench… Live results: benchmarks.coval.ai/tts
Abridge: 100M+ medical conversations, real-time prior auth, and the clinical intelligence layer latent.space/p/abridge @AbridgeHQ is building the clinical intelligence layer for healthcare. In this episode, Janie Lee and @c_asawa explain why ambient documentation was only the first wedge, how Abridge is turning patient conversations into real-time clinical decision support, why healthcare may become one of AI’s most important proving grounds, and how 100M+ medical conversations, specialty-specific evals, and deep EHR integrations create a moat for AI-native healthcare.
