🎙️ "I've been told AI will replace programmers year on year since 2022." That's a comment on the latest episode of my TRXL Podcast with Martyn Day. The person who left it works in AEC from the inside. They know the code. They've watched the hype cycles come and go. You can read the thoughtful (and long) comment yourself over on my youtube channel. They weren't attacking the conversation (I did sense the eye-rolls though 😂). They were asking for middle ground between what gets pushed out into the media field and what most people actually experience on the job every day. That's a fair ask. And a few of their points land cleanly. On the CDE claim: fair. A prototype is not a product. I should have pushed harder on that in the conversation. On tacit knowledge: they're right. AI is limited to what's been recorded. The knowledge that lives in someone's head after 30 years of building, that never got written down, isn't available to train on. There's no software solution to that problem alone. This is very real thing firms have to figure out how to deal with. On timeline skepticism: also fair. The asteroid framing wasn’t about speed. AEC doesn't move fast, and pretending otherwise doesn't help anyone make better decisions. The point is trajectory and whether there's enough lead time to act intentionally. But let’s also apply the premise from the “Don’t Look Up” movie… the gap between firms making deliberate choices now versus firms that aren't is real, even if the pace is slower than anyone on the hype end would have you believe. Where do you sit on this? Is the needle actually moving in your corner of the industry, or is it still mostly noise? What's genuinely stuck? 👉 Episode link in the comments. CC: Gavin Nicholls for some reason or another
Really enjoyed this episode. I won't go into all the points I agree and disagree with. To add to the discussion, I found this article by Radu Gidei a great insight into the tangible impact of AI in a tech-based AEC company; Vibe coding Vs AI agents guided by software engineers. https://www.linkedin.com/pulse/so-whats-changing-radu-gidei-zdqxe/
Look forward to listening to this conversation! Martyn Day's article last month struck a chord with me because we are actively exploring this new world and already seeing some real impact. These are exciting times for the tool builders in our industry! https://aecmag.com/ai/the-ai-wake-up-and-the-death-of-software/
Whenever I hear a claim about how technology [X] is going to do something crazy… I ask the following questions: - Who is making the claim? - Why are they making the claim? (If not made explicit, looking into how they make their living has served me well in the past) - What evidence do they have to back up their claim? (And Is it grounded in something that scales beyond the anecdotal?) A few other key points: - The level of capital investment in a given technology does not guarantee any kind of success or transformative outcome. (Look no further than Meta recently closing the doors on… well…. Metaverse after $80 billion spent.) - High adoption rates of a product claiming productivity gains and transformation does not make transformation inevitable. Look no further than BIM in our own industry where productivity has remained flat or lowered over the period of its adoption. - Productivity gains at a personal level do not automatically scale to organizational wide impacts on metrics like P&L. That’s been a recent headline on studies surrounding AI adoption.
AEC HUB
6d
We are now just building more software to do more things and need more software engineers to help with that. It’s not theoretical I see this in practice: the day to day of a software engineer has already shifted to agentic coding. You will probably notice all the tools you use now have more updates more frequently and new features as well. The problem is many firms don’t know how and where to adapt to best see a return on that investment. And that’s why we think there is a huge opportunity for digital transformation services.
And im only halfway through listening to the episode! Regardless of some of my qualms its a highly valuable discussion - all sides of the future takes deserve and need representation. If it were all grounded grumps like me we wouldn't get far, there's an eventual middle ground this will all settle into I'm sure. Some great points from Nathan there too, agree entirely that the context of claims is so important in a fast consumption media economy where everyone is competing for clicks.
TRXL Media•5K followers
6d👉 https://trxl.co/224