Theo is a developer and founder heavily involved with agentic coding tools, T3 Code, Cursor integrations, and AI model testing. He reviews xAI’s newly released Grok 4.6. He notes xAI has accelerated dramatically since acquiring Cursor, moving from long dry spells to rapid back-to-back releases. Grok 4.6 It is not a new pre-train but a …
November 2026 and Feb 2027 projections by extrapolating the recent doubling rates. Doubling times are taking ~4 months for the amount of human work time AI can perform. Another ~1.5–2× (3 months) to ~2–4× (6 months) increase in effective time horizon is a reasonable baseline expectation if trends hold. This points toward reliable multi-day (or …
SpaceXAI Grok 4.6 is a major release that is by the frontier and competitive with OpenaI GPT 5.6. This is one of the major points on my list of SpaceX Stock Catalysts. Some of the items like Grok 4.7 are adjusted to the new later schedule. SpaceXAI Grok 4.6 is on the terminal bench leaderboard …
Cognition Labs reports Grok 4.6 (available in Devin) is a significant improvement over 4.5, surpassing GPT-5.6 Sol while trailing only Opus 5 and Fable 5. Grok 4.6 (the 1.5T-parameter model with improved supervised fine-tuning and reinforcement learning) launched and is competitive at the frontier. Grok 4.7 is the larger ~2.1T-parameter model. Initial training is complete. …
The AI2027 superAI scenarios are tracking a bit slower but we are still heading quickly to SuperAI. Risk milestones are arriving ahead of the capability milestones that were supposed to produce them. There is a AI2027 tracking website and they observed capability predictions were running a bit behind but the SWE Verified benchmark caught up. …
When you use an AI to write code, you pay for tokens — chunks of text the model reads and writes. Frontier models charge a lot per token. Cheaper models charge a fraction. Think of building a house. An architect charges $500 an hour; a bricklayer charges $40. Both are essential. But you’d go broke …
Cursor has run experiments to test the limits of scaling agents to cooperate toward a goal. Their hypothesis is they can unlock a new tier of task scale and complexity. They are looking to speed up by thousands of times from 1000 commits per hour to 1000 commits per second. Tools like Git and Cargo …
Cursor has built a machine for not needing the best model. The Router sends hard work to frontier models and everything else to efficient ones, delivering frontier quality at roughly 60% lower cost. The swarm research pairs a frontier planner with Composer workers and hits up to 100% on complex tasks like the SQLite rebuild …
Anthropic’s Claude Opus 5 delivers near-frontier performance at roughly half the price, which is a major strategic win for them. This puts competitive pressure on OpenAI, and a clear advance of the practical frontier toward cost-efficient agentic intelligence. It sets new standards for agentic coding, knowledge work, science and business workflows. Metrics overview Opus 5 …
IF Kimi K3 and SpaceXAI Grok 4.5 and OpenAI GPT 5.6 collapse the pricing for Anthropic Fable in the next weeks. Anthropic still has huge realized gains from pushing forward with a 3 month lead for Fable and a 9 month lead for Opus 4.5 and 4.X did let Anthropic capture $30-60 billion in actual …
Gavin Baker says anything that increases competition and compresses margins at the model layer (foundation models) is net positive for every other layer in the AI stack — power, semiconductors, hyperscalers/neoclouds, and even software. A world with only 2–3 dominant high-margin frontier labs is bad for the rest of the ecosystem because those labs become …