Turner released a safety AI tool for free to the whole industry.
SafeT Coach works within ChatGPT and pulls from Turner’s EHS policies, OSHA standards, and years of hands-on experience. Workers can snap a quick photo of a job site condition and speak into their phone to get advice on what they’re looking at. They did 25,000 test interactions before launching it publicly.
Not long ago, Turner rolled out enterprise ChatGPT across their whole organisation without central management, giving users the freedom to create their own tools. This led to the development of over 400 custom AI apps from the ground up, and SafeT Coach was one of them.
In a piece I wrote recently, I pointed to Turner as a great example of this approach: invest in a broad tech layer and let specific solutions grow from it.
The other connection I would like to see with this project is the potential it could generate.
Recently, David Baszucki mentioned that building safety into Roblox isn’t just a compliance hassle; it’s actually a smart way to stand out from the crowd. He pointed out that safety is a compounding factor. Increased trust translates into more users, which in turn yields better data. This whole process not only makes Roblox safer over time but also creates a serious challenge for anyone trying to replicate it.
If Turner sees the same value, making SafeT Coach is the first step to set the standard for what “safety” looks like in a way that machines can understand.
Trace Cohen argued recently that Palantir won by owning the translation layer between how organisations actually operate and how software requires them to operate. That’s the position a widely adopted, free safety AI would occupy in construction.
From building tools for themselves to setting the standard for everyone else. Whether intentional or not, this starts to show a different approach to look at.