An expanding horizon

What’s changing.
What it opens up.

Follow AI’s development through the possibilities it creates for learning, building, and growing.

Our starting baseline · September 7, 2026

These first entries document capabilities available at the time of the source check. They are not release announcements or benchmark results. Future reviewed entries can compare changes against this baseline.

Read the progress data (JSON) ↗
Building · Capability snapshot

From code suggestions to delegated tasks

Observed 2026-09-07

What the source says

Cursor describes agents that work on software tasks and produce changes for review.

The opportunity

You can explore a small software idea by defining the outcome, then reviewing and testing the work an agent produces.

What remains uncertain

Delegation still requires clear requirements and verification. This entry does not establish a measured productivity gain.

Cursor product overview ↗ · Try a related guide

Creating · Capability snapshot

An outline can become a visual first draft

Observed 2026-09-07

What the source says

Gamma advertises generation of presentations, documents, websites, and graphics from prompts or existing content.

The opportunity

An idea can take a shareable form earlier, giving you something concrete to refine and discuss.

What remains uncertain

A generated draft still needs factual, visual, and accessibility review. This is an available capability, not a newly verified release.

Gamma product overview ↗

Learning · Capability snapshot

A conversational starting point for difficult work

Observed 2026-09-07

What the source says

Claude presents learning, coding, research, analysis, and creation as supported tasks.

The opportunity

You can ask for explanations, develop an initial plan, or examine a rough idea before deciding what to do next.

What remains uncertain

Useful dialogue is not evidence that every answer is correct. Check claims against reliable sources.

Claude product overview ↗ · Try a related guide

What we’ll look for

New capabilities, easier access, changes in cost, and evidence of better results on real tasks. Each report should separate the vendor’s claim from independent evidence and our interpretation.

Updates in this version are reviewed and published manually. There is no live feed or automated reporting running.