diff --git a/docs/src/SUMMARY.md b/docs/src/SUMMARY.md
index b035903772..f0131dc751 100644
--- a/docs/src/SUMMARY.md
+++ b/docs/src/SUMMARY.md
@@ -31,6 +31,7 @@
 - [Terminology](reference/terminology.md)
 - [Common pitfalls](reference/common_pitfalls.md)
 - [FAQ](reference/faq.md)
+- [The path to 1.0](reference/path_to_1.0.md)
 - [Changelog](reference/changelog.md)
 
 # Runtime internals
diff --git a/docs/src/reference/path_to_1.0.md b/docs/src/reference/path_to_1.0.md
new file mode 100644
index 0000000000..c8a7da0239
--- /dev/null
+++ b/docs/src/reference/path_to_1.0.md
@@ -0,0 +1,24 @@
+# The path to 1.0
+
+The development of Ambient can be split up into two rough periods; pre-1.0 and post-1.0.
+
+Pre-1.0, which we're currently in, involves heavy iterations on the API and services. This
+is a period where we experiment with things, where we take big and wide decisions, and where
+we're not afraid to throw things out when we find better ways of doing things. During this
+period you can expect fairly frequent breaking changes to the API and services. Very little
+effort will be put into maintaining backwards compability.
+
+Post-1.0 this will reverse. After 1.0 we will strive to maintain maximum backwards compability,
+and very rarely introduce breaking changes. In the post 1.0 era you should expect much greater
+long-term stability, and you should be able to trust that what you build will work for as long
+as you're using Ambient as a service.
+
+So when will we reach 1.0? These are the rough things that we think need to be completed before
+we can do that:
+
+- [ ] Guest side programmable graphics (user shaders)
+- [ ] Performance roundup; we should take one big last look at how api's are structured to figure
+  out if there are any broad sweeping changes we need to make to maximize performance
+- [ ] One at least medium sized game needs to have been built with the API, to work out the worst kinks
+
+Once those things are in place, we'll set a date for the 1.0.