Why old AI product ideas like Rewind and Alexa may work now
marvinvonhagen · x · 2026-07-29
Old product ideas are getting a second chance now that agents and voice models exist
The post argues that many “ahead of their time” products are becoming viable again because the underlying AI stack has changed.
- Rewind is cited as an example: four years ago, there was no Claude or Codex to make sense of terabytes of screen-recording context. Today that kind of memory layer looks much more useful.
- AutoGPT is framed as an early attempt at LLMs that could take actions and use tools, but the form factor only became mature later.
- Alexa is mentioned as another case of arriving before real-time, full-duplex voice AI existed.
- The author points to a new project, AttentionInc, as an example of the “why now?” thesis: old ideas can work if the model capabilities and interfaces have finally caught up.
- The post ends by asking which other “old” product ideas could now succeed with today’s AI stack.
More from Companies & People
- Apache Airflow’s AGENTS.md says agents are assistants, not commit authors — eyishazyer · 2026-07-29
- Open-weight AI still follows an old business playbook, Kevin Xu says — kevinsxu · 2026-07-29
- PRIME BPM adds four AI agents to automate process design and documentation — Free-Explanation-124 · 2026-07-29
- Walden Robotics teams up with Samsung SDS on factory robot pilot — RussTedrake · 2026-07-29
- Zuckerberg Aims to Starve Closed AI Labs with a Price War — Longjumping_Cold7410 · 2026-07-29
- OpenWiki tops 23.4K weekly downloads as an agent wiki CLI for codebases — LangChain · 2026-07-29