Directories of organisations and activities is always a challenge.
Working with Nautoguide, they have used AI to collate and check a national list of sports clubs to refresh and update our directory.
In doing so it has created a national list, and we see the value to integrate all this this with systems using Openactive and Open Referral - especially with the role sport can play in the new ‘prevention’ shift, AI has the patience to keep databases updated so social prescribers know where activities might take place.
On another topic, an ICB is building a tracker to coordinate public health campaigns, which would involve a directory on NHS Futures site - comms people did not seem aware of Open Referral - so I showed them this site.
Thanks @Julz. We have an NHS standards person on the ORUK Cross Government Steering Group and one ICB (Lancs and South Cumbria) is adopting ORUK enthusiastically. I’d welcome more details.
Are the websites for sports clubs generally kept reliably up to date? If so, then I suppose AI tools might be a plausible way to gather this information at scale.
Certainly in the human service and community resource sectors, the root of our problem is that organizations don’t reliably update their own websites with clear information. So you get GIGO, and then that is amplified by most AI tools’ inherently chaotic natures . . .
But it would still be great to learn about other fields where the incentives and information processes are inherently more aligned.
VCSE is massively dependent on the work of goodwill, but the internet is the best we have.
Most records are from NGB and other sources like Sport Engand’s Active Lives so we are bit more confident.
I am a big fan off AI despite literally spending 50 years reading and watching it going spectacularly wrong.
Usually by not understanding the question.
In this case, I’ll just reiterate the main lesson from decades of this work, which is that in this domain the real question is about how to establish designated, sustainable human capacity to produce reliable data. AI magic carpet rides can happen downstream from that. Without solving the (human) data supply chain problem, the robots might only make things worse.
