Thoughtworks Technology Podcast · Thoughtworks

Be brilliant at the basics: Inside Looking Glass 2026

·46 min·2 clips

Thomas Leonard says AI ROI is not hours saved or lines of code, but business outcome and risk.

The report becomes a field problem. Ken frames Looking Glass through what clients and Thoughtworks teams now have to do with it, then brings in Thomas and Ricky because their work sits close to the agentic and platform shifts in the report. February 2026 gives the episode its tempo. Thomas calls it a significant inflection month for what Thoughtworks has been seeing in AI, so the conversation has a practical sense of things speeding up without turning breathless. Measurement is the center. A manual spreadsheet testing example helps separate real automation uplift from the looser talk around productivity or time saved. Hours saved do not carry the argument on their own. AI value has to show up in leading indicators tied to the capability being changed, such as automated tests or tests using AI elements. DORA and SPACE still matter, but the hosts treat them as part of the toolkit rather than a finished answer for AI-enabled or AI-native engineering. Thoughtworks is still researching what the next set of metrics should look like once token costs, developer behavior, automation, and value start pushing on each other. No neat answer appears, which is part of the point. ROI depends on context. Jevons paradox gets a nod too, because freed capacity may simply turn into more work instead of clean savings. The CTO role widens as AI leaves the CTO-only corner and reaches stakeholders across the business. Technical leaders now have to teach enough fluency for others to judge capability and consequence. The closing thought has teeth: for a CTO, having no AI opinion is becoming a short-lived position.

As heard by us

A grounded look at AI adoption, measurement, and the CTO role as the conversation moves toward enterprise fluency.

ThoughtWorks uses its Looking Glass report to open a practical discussion about how AI is landing in client work, platform engineering, and the companys own agentic journey.

Read the full review in PlayNext →

Why you'd press play

If you need sober AI adoption metrics, this one stays grounded in production reality.

Read the full recommendation in PlayNext →

Listen to the show on

Apple Podcasts Spotify

Play clips from this episode More from Thoughtworks Technology Podcast