Episode 170 of the podcast is available now via this link or from the usual podcast platforms.
"The second (epochal change) is the technology revolution led by developments in artificial intelligence, which will change everything. I mean everything. There is no point in debating whether this technological revolution is a good or bad thing. Just know it is a ‘thing’. In fact, it is ‘the thing’. It will displace jobs, though creating new ones, but no one yet knows the full consequence. Companies and countries will rise or fall on the back of it. It will revolutionise the private sector and should in time revolutionise public services and government. Yet people in most countries, including Britain, have no idea what is about to hit them."
Not my words, Readers, but the words of Tony Blair in his recent essay.
What is about to hit us? What are the implications of AI in the planning context? What does it mean for what we do now and what we might do in the future? What are the legal implications? The data implications? The implications for public engagement?
These themes were explored in a conversation recorded online back in February of this year between old friends of the podcast Hashi Mohamed, Kathryn Ventham and Sue Chadwick, and new friend of the podcast Harry Quartermain.
Hashi, who steered the conversation, is an author and a barrister at Landmark Chambers; Kathryn is a Senior Director at Twenty5 Planning; and Sue is a strategic advisor at Pinsent Masons. Harry is Head of Insight & Research at LandTech and an Associate Lecturer at the University of the Built Environment.
Due to some technical issues on the day another new friend of the podcast, Graham Stallwood, was unable to join the recording, but Graham, Interim Chief Executive at the Planning Inspectorate, has since listened to it and kindly provided an addendum that I have dropped in at the end.
Subsequent to that recording, somebody, who I assume was preparing to speak at a conference, asked the following question in a WhatsApp group to which I am party: ‘If you were to talk about the three main issues around AI and planning what would they be?’
It prompted a thread that was so good that I asked the contributors if I could capture it. The consensus on those three main things was as following.
Data
AI is only as good as the data it uses, and the consensus seems to be that not only are we starting from a low baseline, digital maturity within LPAs is low as well. Plan-making, at all levels, offers the opportunity for a reset, but will that opportunity be taken?
Efficiencies
Within the plan-making process itself there might be, for example, opportunities in relation to identifying sites and their capacity. The greater opportunities to free up officer time within LPAs could though be using AI for PD and householder applications, many of which are more algorithmic.
Objections & Engagement
Any discussion about AI and planning either starts with or ends with the risks of AI-based objections and inadequately verified evidence.
The tide of AI-generated objections risks gumming up the system quickly if action is not taken soon. Extending the PINS processes to applications by way of Planning Practice Guidance would be a quick win and other initiatives might include the setting of mandatory word limits.
We should perhaps though take a moment to step back because in the rush to deal with the AI objection issue we must not lose sight of the public’s right to participate and comment.
How though? AI is changing how the public interact with services, which makes the need for clear and accessible information even more important.
Why though? Let us go back to first principles. What are we hoping to gain from public participation and what process best achieves that? We are currently trying to adapt an existing system to new technologies without considering what we hope to achieve by collecting comments in the first place.
Hundreds of people might be willing to express a view on an application on a Facebook, but only a few might submit a formal comment? Does that matter? If it does how can those views be captured in a worthwhile way?
By and large people never 100% like or 100% hate something, but if asked a binary question (like support or object) they are perhaps more likely to object to something because of a tiny part they dislike on the basis that they do not support it enough to endorse it completely. How about a five point Likert scale?
What though if the principle of development has already been established and the proposal is fully policy compliant? Does the public have a right to participate and comment then? Who is going to tell them if not?
All of the risks associated with AI arguably exacerbate issues known already so does the profession need a livelier debate about ethics and reasonable expectations? If we cannot set the tone with it be the tech companies that start to answer questions like acceptable limits to community input and how rules-based should planning decisions be.
These three issues are though very much of the here and now. They are but part of a wider discussion about our 'ways of working'.
Cleverer people than me are saying that the advent of the technology behind AI is as significant a shift as was seen during the industrial revolution. Just like then, it is going to disrupt where, how, and on-what people work.
At the minute there are arguably two planning systems: planning and digital planning. For how much longer will we be 'digitising' the current (paper-based) system when what we could, should, need to be doing now is 'digitalising' the aims and objectives of the system using the best technology available.
The AI we have now, so I am told, is the worst it will ever be. I is reportedly improving every couple of months, and quantum computing - which I heard being discussed on the Today programme recently is predicted to be less than a decade away from a commercial application - will exponentially boost the capabilities.
Extract is turning legacy pdf policy into shape files. New SDSs and local plans will provide digitally native policy layers as well. New legislation and the NPPF is 'rules based', which is step more towards code than subjective policy.
As Harry has said to me, what we are witnessing now is just the tip of the AIceberg...
Please direct all complaints about the title of this episode to him...
Comments
Post a Comment