Capabilities that have crossed the line. Oldest flip first — a chronology of the frontier.
Stylometric modeling is good enough now that period-correct pastiche is generated on demand. Pianists actually play these. --- AI systems have demonstrated th...
Models that combine social signal, trailer engagement and historical patterns now beat box-office veterans on aggregate predictions. --- AI systems have made...
A capability we wish AI didn't have. The defensive flip side — AI phishing detectors — also got dramatically better in 2024. --- AI systems can generate text...
Modern music-LLMs handle counterpoint, voice independence, and harmonic propriety. Not Bach himself, but credibly Bach-inspired. --- AI systems have demonstra...
Specialised diffusion models for character rigging shipped in 2024 and immediately changed the indie game animation pipeline. --- AI can generate animated cha...
Order tracking, returns, FAQ — entirely automated tier-one support. Humans now mostly handle escalations. --- AI can run a customer-support chat unsupervised...
Waymo and Cruise expanded to dense urban areas (with setbacks). The bar for 'safe' continues to be argued, but the technical capability is evident. --- Curren...
Diagnostic-companion models in 2024 found cases of rare conditions missed by clinicians in both training data and live trials. --- AI can diagnose certain rar...
NDAs, employment offers, lease agreements. Big-law associates spend more time editing than drafting now. --- AI can generate court-ready first-draft contracts...
Continuous sign-language translation crossed the usable-prototype line in 2024. Not deployed broadly yet, but demonstrably working. --- Researchers have made...
State by state, including the road-sign questions and edge-case rules. Trivial for any modern frontier LLM. --- AI systems have made significant progress in n...
GitHub Copilot Workspace, Sourcegraph Cody, others — most modern engineering teams use AI-generated review comments as a first pass. --- AI can generate code...
Cursor + SwiftUI generation hit a quality bar in 2024 where solo founders ship App Store apps without writing a line of native code. --- Current AI technology...
Most major IDEs now suggest tests automatically from function signatures and docstrings. --- AI can generate working unit tests from a description of intent t...
Tools like FunSearch and AI-co-scientist released in 2024 surfaced novel hypotheses in materials science and biology that humans then verified in lab. --- Cur...
Specialised math models plus chain-of-thought tooling closed the gap to top human contestants in 2024. --- AI systems have demonstrated the ability to perform...
Runway, Pika and Sora's full release in late 2024 made full-shot generation a routine indie-film workflow. --- AI can generate animated short scenes from a sc...
Apple's translation earbuds, Google's Pixel Buds Pro 2, Meta's Ray-Ban — speech-to-speech translation became a consumer feature in 2024. --- Current AI system...
AlphaFold-Multimer and successors took this benchmark in 2024. --- Current AI systems have made significant progress in predicting protein-protein interaction...
Research models picked up Alzheimer's markers in 30-second voice recordings years before clinical diagnosis. Not yet a routine clinical tool. --- Researchers...
Two-host banter, ad reads, transitions, 30+ minute show length. Many indie podcasts are now mostly LLM-scripted. --- Current AI systems can generate text base...
AMP Robotics and competitors automated the dirtiest job in waste management. Better than the average sorter, runs 24/7. --- AI systems can sort recyclables on...
Research-grade tools, mostly used in screening and not as standalone diagnoses. Effective enough that several universities pilot them in counseling intake. ---...
The 'aha' moment problems that used to stump LLMs are now mostly solvable with good chain-of-thought tooling. --- AI systems have made significant progress in...
Add a statement to the atlas. We review weekly.
Site
stuffaicantdo.com © 2026 · made in NL by Arcadist
Pick a side, give it a one-line title, optional explanation, and a category. We review submissions weekly.