Expert Insights

Alan Buxton discusses a major shift in AI-assisted software development, naming it "The Speed Inversion," where the speed of development now outpaces the speed of discovery. He sheds light on the new challenges and changes that this reversal of pace has brought to the software industry.

Alan stresses the new reality:

  • The shift in question has moved from "can we build it" to "should we build it," alluding to the necessity of discerning what is valuable to build in the first place.
  • Identifies a significant issue most organizations face; their structure still operates as if development is a time-consuming process.
  • Instead of the traditional focus on coding, he emphasizes the importance of rapid experimentation, strategic product thinking, and judgment.
  • He warns about the missteps by organizations in overestimating how much AI can take up and underestimating how much of that they will actually utilize.

Quote

quotation-marks icon
So in the olden days, you'd have written it down on a whiteboard and then you'd have got someone to do some figs and then you'd have built a prototype. Then you'd realize that wasn't really what you wanted... anyway, so he spent two hours vibe coding, something. All this stuff that used to be quite painful and boring, like upgrading Python from two to three or migrating from one tech stack to another... All the stuff that would take you days of time is now pretty much instant. I am optimistic about the things that it's good at and I'm pessimistic about the state of the industry at figuring out what it's good at.quotation-marks icon
Alan Buxton ,
CTO, Simphony

THE NEW DEFAULT angle

Here are practical steps to consider to capitalize on the speed inversion in AI-assisted software development:

  • Adopt a more discerning approach to ideation: Acknowledge that speed of development has surpassed speed of discovery, and shift your questions from "can we build it?" to "should we build it?".
  • Redefine team competence: Traditionally, coding has been the core skill for software teams. With AI, strategic product thinking, rapid experimentation, and sound judgment become crucial.
  • Educate your teams on AI capabilities and limitations: Knowing when AI is applicable and efficient is equally as important as knowing when it isn't. Train your team to avoid misuse or overuse of AI.
  • Embrace a culture of rapid testing: Encourage your team to use AI's speed for validating a variety of ideas rapidly, learning from the outcomes, and iterating smarter.
  • Reexamine your team structure and processes: Traditional sequential workflows and meticulous planning may not serve in the era of AI. Transition towards parallel workflows and rapid testing.
  • Use AI to automate repetitive tasks: Actions such as code optimization, unit testing, and tech stack migration can now be handled by AI, freeing up your team for more strategic tasks.
  • Prioritize learning and adaptability: Given the increasing pace of development, the ability to learn and adapt quickly becomes crucial. Reinforce a culture that rewards these traits.