Capabilities
On-Device AI: What Your Phone Can Actually Run Locally
A breakdown of what on-device AI models can handle on a modern smartphone, and where they still hand off to the cloud.
Capabilities
13 articlesWhat AI systems can actually do, tested against the claims.
13 articles
-
-
CapabilitiesWhy AI Coding Assistants Still Struggle With Legacy Codebases
Legacy codebases break the assumptions most AI coding assistants were trained on, and the gap shows up as soon as you leave a clean, modern repo.
-
CapabilitiesAI Agents vs. AI Assistants: What's Actually Different
Vendors call almost everything an agent now. Here's the technical line that actually separates an AI agent from an AI assistant, and how to test which one you're using.
-
CapabilitiesLong Context Windows: What a Million Tokens Buys You
A million tokens holds 700,000 words. Attention across them is uneven, and you pay for every one. Where long context wins and where retrieval still does.
-
CapabilitiesAI Coding Agents in 2026: What They Finish Unattended
Tests, refactors, and reproducible bugs go well. Vague specs and unwritten conventions do not. What changes when review becomes the bottleneck.
-
CapabilitiesAI Benchmarks Explained: What the Scores Actually Mean
Contamination, saturation, prompt sensitivity, and averaging. Four reasons a leaderboard number tells you less than ten minutes with the test items.
-
CapabilitiesAI Code Review: Where Automated Review Actually Catches Bugs
AI code review tools genuinely catch real bugs before they ship, but they’re reliable for specific categories of problems and weak for others worth knowing in advance.
-
CapabilitiesAI Video Generation: How Close to Real Is It Now
AI video generation has moved from a few seconds of shaky, dreamlike motion to longer, more coherent clips — but real limitations still separate it from footage.
-
CapabilitiesAI Image Generation in 2026: What’s Actually Improved
AI image generators have gotten dramatically better at text rendering, hands, and following detailed instructions, though specific weaknesses remain.
-
CapabilitiesAI Hallucinations Explained: Why Models Still Make Things Up
AI hallucinations aren’t random glitches — they’re a predictable byproduct of how language models generate text, and understanding why helps you catch them.
-
CapabilitiesCan AI Models Do Real Math? Where They Still Fail
AI models have gotten dramatically better at math, but the specific ways they still fail reveal a lot about how they actually process numbers.
-
CapabilitiesAI Voice Cloning: How Realistic Synthetic Speech Has Gotten
AI voice cloning can now recreate a specific person’s voice from just seconds of audio — a capability that’s genuinely useful and genuinely easy to misuse.