Organisational strategy for AI-Assisted coding adoption
The world of AI-assisted coding is fast-moving, changing and evolving. Every week, new models, tools, and IDEs are released, making every knowledge and practice temporary.
The entire field of AI-assisted coding itself is in its infancy. Our understanding of how to use AI coding assistants effectively is constantly evolving. There is much more to discover and learn, than what we know already. What we know today will become obsolete relatively quickly.
The only certainty in this phase is change. For organisations that crave certainty and stability, and control, this can be hard.
The biggest pitfall an organisation can make at this stage includes:
- Prematurely committing to AI tools, models, vendors (like for example annual enterprise subscriptions)
- Defining organisational AI standards and best practices (the best in the game are still experimenting and searching for answers)
- Falling for meaningless AI-Assisted coding Certifications (at this stage there is no curriculum worth the paper it is written on)
A much better alternative
A better strategy for an organisation to adopt AI coding successfully, requires the Tech leaders empowering their teams to:
- Experiment Continuously: give teams the mandate to continuously try out the new and emerging AI tools and models, providing the authority and budget (enabling them to quickly buy monthly subscriptions, install and use them)
- Explore New Ways of Working: encourage every team to explore and experiment with new ways to get the best out of LLMs and AI augmented IDEs and tools (to revolutionise the organisation’s ways of working)
- Systemise Knowledge Sharing: enstablish a system for coordination and knowledge sharing around these experiments, both within and between teams. This creates a critical debate, leading to new ideas and accelerating collective learning.
This inward-looking approach should be balanced with a list of external sources to follow, for inspiration.
The organisation should leverage a tool like the ThoughtWorks Tech Radar to visualise the models, tools and IDEs, as well as the technique of work, and their stage (Assess => Trial => Adopt / Hold), as a visual support to organise and coordinate the experimentation among teams, and track the progress.
Something also for the leaders
This period of exponentially accelerating change and innovation may feel unfamiliar to many organisations and leaders. The strategy outlined above may challenge their usual operating models..
For a deeper, more methodical view of this approach I strongly recommend this reading => The forgotten new philosophy of work, management & leadership
Develop Technical Excellence that delivers.
See how we can help.
You, your team, your Tech.