Reliable custom software & support in plain English
The distinction between ‘describing’ and ‘understanding’ really hits home. We’ve been so focused on chatbots that can talk, but VLJ’s approach to actually comprehending temporal meaning feels like the missing piece for real-world AI. Descriptions up from atomic level is absurd for nearly everything, yet that is precisely how LLMs presently operate. Abstraction is key. … Read more
Docker pioneered but promptly lost control of containerization. I see the same happening to OpenAI / ChatGPT. “Eventually openai will be bought by Microsoft and that’ll be that.” “Cancer [cure] is 5% baking soda in water as revealed by Dr Simoncini (as ancient cure) before he was run aground…” “Data Center apocalypse in Northern Virginia … Read more
LLMs as we know them are already on their way out. In this video, I break down five breakthroughs that will redefine AI over the next 18 months: Diffusion LLMs (with Stanford’s Stefano Ermon), Power Attention for massive context, hidden/latent-space thinking and private chains of thought, Google’s Nested Learning for continual learning, and the most … Read more
80% speed gain via Pre-Aggregated subsets vs MongoDB (which gains on bloat) On PostgreSQL https://medium.com/@kanishks772/we-cut-80-of-our-query-time-by-using-this-little-known-sql-pattern-fadec2bdb592 Compare against MongoDB To precisely compare real-world speed between PostgreSQL pre-aggregation using CTEs vs MongoDB aggregation pipelines — Below is a detailed side-by-side analysis — MongoDB 10% slower with smaller datasets — But MongoDB speed advantage increases with increasing datasets … Read more