Aggregate stats often result in misleading conclusions. Here’s a good example from @a16z ( that implies Uber’s prices in the US are up significantly more than Lyft. At Uber, we measure this particular metric every single week. But to get to the true signal, we always control for mix effects from trip type and geography. We measure price quotes on a standard product (in this case UberX vs. Lyft Standard) on identical trips (same origin/destination/time of day combinations), and we call this metric NPI (normalized price index). Our philosophy consistently has been to price as affordably and as competitively as possible, with the NPI near 1.0 at all times. On an apples to apples basis, that story remains remarkably consistent in the US. Even though the overall assessment is directionally accurate, in not controlling for mix effects, A16Z/Gridwise come to an inaccurate conclusion. In fact, it shows the success of our deliberate strategy to both innovate with new products and expand our footprint into sparser markets. As a result, (1) Uber skews more toward higher-priced premium products, as consumers are increasingly opting into our newer premium products, and (2) Uber trips skew longer, and the mix of trips >15km has increased meaningfully driven by the faster growth in suburban and sparser markets. In other words, we are serving consumers with better products, and we are serving more consumers than ever before in newer markets that have been significantly underpenetrated for the ridehailing category.

April w/software down 50%: integration of s/w with Claude is fox in the henhouse. Dumb pipe. It’s over. August w/software up 100% from the lows: Claudeforce is a step function change. I use okta to identify all of my wife’s bfs as they tagTEAM her. Buy. Bunch of inbred retards

$SNOW continues to lose market share, and it's not just Databricks they need to worry about. While the entire market is growing based on hiring demand, the number of companies mentioning $MSFT Fabric jumped the most at +165%. Databricks grew +40%, while both $SNOW and $GOOG BigQuery increased the least at +15% from a year ago.

Solid work by our teams to bend the AI cost curve while continuing to enable broad adoption. AI adoption at Uber is up 7X since Feb, while cost per session is down over 50% since the peak in May. More here:

Good luck to Meta on its new journey with the new settlement/time limits for teens. When Australia banned all social media for kids under 16, it was effective for like... 3 weeks (and then kids immediately got around it). My story:

Codex B2B adoption continues to accelerate while Claude Code adoption has rolled over. Over the last 3 months: Codex: +183 pts Claude Code: -157 pts A sharp divergence in enterprise adoption trends between OpenAI and Anthropic.

Guide is $420M rev, $20M ebitda $104M net debt $21M mkt cap $125M ev Equity stub 5-6x ebitda Rev -6.5% but showrooms growing 17% Might dust this turd off again but rates aren’t helping w/housing turnover