nthypes

After selling to mid-market customers, I've become convinced that multi-tenant SaaS architecture is the wrong model for anything beyond PLG at low price points. Enterprise procurement (and some mid-market) wants…

I've been thinking about how vibrant web forums used to be in the 2000s and early 2010s - specialized communities with deep expertise, long-running threads, inside jokes, and real relationships. Places like Something…

I've been trying out Google's new AI Mode in Search and I'm genuinely curious about the technical architecture behind it. The response times are incredibly fast - often sub-second - and the quality of answers seems…

Last month, a single container in my GKE cluster (Sao Paulo region) entered an error loop, outputting to stdout at ~2k logs/second. I discovered the hard way that GKE's default behavior is to ingest 100% of this into…

I've asked Sonnet 4.5 (Cursor) to generate some stats based on my cursor CSV file usage. Here it is the output: Key Findings from Your Cursor Usage Here are the most interesting insights from your usage data: Cost…

I just got my Cursor invoices and I'm genuinely shocked. Between October and early November, I've burned through $638 on AI coding assistance. That's more than some cloud bills. Context: I'm an Founder & CTO building an…

We have the monthly "Who is Hiring?" and "Who Wants to be Hired?" threads, which are incredibly valuable for connecting talent with opportunity. Let's experiment with the other side of the marketplace: a thread for…

In various fields, we often encounter the principle of Sturgeon's Law, which suggests that "90% of everything is crud." When it comes to datasets, how much of this holds true? With the proliferation of LLM are we seeing…