13 companies · 7 sourced backers · 87 milestones on record
This frontier covers broadly capable AI models and the data, cloud services and computing chips needed to build and run them. Companies aim to make these systems useful for everyday work, software development, science and manufacturing, while improving safety, human control and the ability to adapt models to specific tasks.
Where the 13 companies stand today, from their profiles and 10-year milestones. Tap a phase to see who is there.
- ▸Making powerful models safe, understandable and controllable
- ▸Turning model capabilities into dependable applications
- ▸Providing the data and computing infrastructure models need
- ▸Adapting general models to specific tasks
Publicly listed: 6 companies (owned by public shareholders).
- ›Every sourced story on all frontiers, filterable by company and news type
- ›10-year milestone records for each company: demo vs shipped
- ›Explained: plain-words readiness diagrams for real breakthroughs
- ›Signal Trends of 39 frontiers covering 378 companies over time.
- ›The events calendar: launches, approvals and deadlines ahead
- Alphabet (Google DeepMind)
To build general, highly capable AI and apply it both across Google's products and to major scientific problems, running it largely on Google's own custom ch…
Everyone · publicly listed
- Amazon (AWS AI)
To make advanced AI practical for organisations through AWS cloud services, model access and purpose-built computing chips.
Everyone · publicly listed
- Anthropic
To build frontier AI systems that are highly capable and steerable while remaining safe and interpretable, so that increasingly powerful AI can be deployed w…
Everyone · publicly listed
- Inception
To make diffusion-based language models a practical alternative for generating code and text.
Some users · backed by Disruptive
- Meta Platforms
To develop increasingly capable AI models and make them useful across products and developer applications.
Everyone · publicly listed
- Microsoft
To embed AI assistants ('Copilots') across its software and cloud, and to run them on Microsoft's own optimised hardware, so that AI becomes a standard layer…
Everyone · publicly listed
- OpenAI
OpenAI's stated mission is to ensure that artificial general intelligence (AGI) — AI that is broadly smarter than humans at most economically valuable work —…
Everyone · backed by SoftBank
- Prometheus
To apply AI directly to industrial engineering and manufacturing.
Lab
- Reflection AI
To develop capable open AI models and autonomous coding systems.
Some users · backed by Nvidia
- Safe Superintelligence (SSI)
To develop superintelligent AI while making safety a central property of the system.
Lab · backed by Nvidia
- Scale AI
To provide the data and supporting infrastructure needed to develop useful AI models.
Some users · backed by Meta
- Thinking Machines Lab
To make frontier AI more understandable, customisable and collaborative.
Lab
- xAI
To develop increasingly capable AI models and make them useful through Grok and related products.
Everyone · publicly listed
Keep following breakthroughs like these. Every Monday, get up to 5 major breakthroughs and launches explained in plain English.
Free. No account needed. One email a week. Unsubscribe anytime.