Eleven things on Technology Radar Vol.33 that aren’t at all related to AI

Recent editions of the Thoughtworks Technology Radar have reflected the increasing adoption and impact of AI in every part of the software industry. The most recent volume in particular was dominated by AI — around two thirds of the 114 blips were AI-related. That’s an important signal of the technology’s importance today, but it would be wrong to think AI is the only game in town. There are lots of interesting things happening in many parts of the industry and some exciting innovation which isn’t AI-related at all. 

So, with that in mind we’ve selected a handful of things you may have missed on the most recent edition of the Technology Radar that are absolutely not related to AI in any way. Of course, it’s worth having a disclaimer: it might well be possible that any of the things featured below could, at some point, be related to AI, but, right now at least, they’re not part of that conversation. If you disagree, please send your responses on a postcard…

With that out of the way let’s get to it. Here are 11 interesting things on Technology Radar Vol.33 that aren’t related to AI at all.

Development frameworks worth exploring

This edition of the Technology Radar includes a number of interesting development frameworks. Notably, Fastify (Languages & Frameworks/Adopt), a lightweight and minimal web framework for Node.js, was moved into Adopt because we’ve seen consistent success with it. (Previously it featured in Trial five years ago!) As we note in the report itself, we can see no downsides when compared with alternative frameworks like Express.js. 

Also featured in Vol.33 is Nuxt (Languages & Frameworks/Trial), “an opinionated meta-framework built on top of Vue.js”. It’s particularly useful for building SEO-optimized websites thanks to the way it manages pre-rendering and server-side rendering, and also integrates very nicely with a number of modules, supported by Vercel. 

Another framework worth mentioning is Phoenix (Languages & Frameworks/Trial). Built on Elixir, Phoenix isn’t a new framework — in fact, it featured on the Technology Radar in March 2017. However, we wanted to flag it again because of its new LiveComponent feature. LiveComponents are, as the documentation states, “a mechanism to compartmentalize state, markup, and events for sharing across LiveView”. We think this is pretty neat as it helps developers build interactive frontends without a full, heavyweight JavaScript framework.

Finally, we’d also like to highlight Tauri (Languages & Frameworks/Trial). Tauri helps teams build desktop applications with a single web UI codebase. It featured on the Radar in October 2021 (in Assess), but we wanted to include it once again — this time in Trial — because the latest version comes with some updates we like. These include a more flexible permission and scope model and a hardened inter-process communication layer.

Developer tools

If you’ve explored the Radar already you’ll know there are lots of AI-related coding tools. However, to pretend that everything is now AI would be wrong; there are plenty of non-AI tools featured we think you should explore. 

For instance, we like the API development tool Hoppscotch (Tools/Trial). It’s open source and also exceptionally lightweight; its privacy-friendly design is great too — it doesn’t collect analytics and uses local-first storage as well as supporting self-hosting. We also featured Helix (Tools/Assess), a ‘post-modern text editor’ according to the team behind it. It’s part of the current wave of editors seeking to replace Vim; we appreciate its support for Tree-sitter and Language Server Protocol (LSP).

We should also mention pnpm (Tools/Adopt), a Node.js package manager we’ve seen good results with. We think it offers significant performance benefits over other options, such as Yarn.

Infrastructure and cloud

There were many blips in the cloud and infrastructure space that weren’t related to AI. Oxide (Platforms/Assess) is one particularly interesting blip. It provides prebuilt racks with compute, networking and storage, and runs fully integrated system software. It enables what the company describes as “on-premises elastic infrastructure.”

We should also highlight Uncloud (Platforms/Assess), a container orchestration and clustering tool that helps you take Docker Compose applications to production without the operational overhead of Kubernetes. We note in the publication itself that its “main architectural advantage is its fully decentralized design, which eliminates the need for a central control plane and ensures cluster operations remain functional even if individual machines go offline.”

There were other blips that were inside or at least touched by Kubernetes. Crossplane (Languages & Frameworks/Trial), is one worth noting in particular. It’s a framework for extending K8s clusters. We’ve seen some instances of Crossplane being used as a full replacement for Terraform, but we’ve found it to be most successful as a companion to workloads deployed within Kubernetes. 

Finally, we wanted to call out one technique. Almost every technique in the latest volume was related to AI in some way, but service mesh without sidecar (Techniques/Assess) is, we think, an interesting anomaly. It appeared on the Radar back in 2022, but we thought it was worth featuring again as Istio Ambient Mode is proving a very effective option for such an approach. 

To be clear, this isn’t an exhaustive list of things on Technology Radar Vol.33 — there’s many more. Nonetheless, this is an important reminder that although AI may dominate conversations and impacts just about everything in some way, there’s still much more to the modern software industry than AI.