In "Crime: Numbers, Narratives and Nuance" our host Miles Fletcher speaks with Nick Stripe, Joint Head of Crime Statistics at the ONS and John Rentoul, leading commentator on crime, policing, and the media, about the challenges in interpreting crime data.

Transcript

MILES FLETCHER

Hello and welcome to another episode of Statistically Speaking – the official podcast on the UK's Office for National Statistics. The time we're returning to the scene of a major statistical topic we've touched on before but amid a new and sometimes highly polarised public debate, one we think fully bears further investigation: how best to understand and interpret the crime figures produced and published by ONS.

Helping us with our enquiries is Nick Stripe, Joint Head of Crime Statistics at the ONS. It's his job to assemble and present the complex statistical picture of crime revealed in two very large and sometimes conflicting data sources. We also have an independent witness in the highly experiences shape of John Rentoul, Chief Political Commentator for the Independent and visiting professor at Kings College London. He'll be talking about the use and possible abuse of crime figures in the media and political debate. Are the statistics and those who produce them doing enough to enable the public to understand properly the prevalence and nature of crime in our society today?

Nick, a big question to start with, some people think crime is going down, other people insist it's going up. Who's right?

NICK STRIPE

Well, it's a question that sounds simple, doesn't it? And it's a question I get asked quite a lot. But if you think about the concept of crime, you soon realize that it covers a really huge range of actions and behaviours. If I was a chief constable trying to reduce crime in my area, I'd want to know what kinds of crimes are causing the biggest problems. So, is it theft, robbery, violence? Domestic abuse? sexual offenses? Maybe it's fraud. And even if you tell me it's theft, then there's still a broad spectrum. So, is that burglary from houses? Is it theft of vehicles? Is it people having things snatched off them in the street? Is there a new thing about theft from doorsteps? Each of those types of theft would have its own trends, patterns and challenges. So, what I'm really saying is that whilst I understand your desire for a single answer, the real stories are in the detail of those different crime types.

But I will come back to your question, is crime going up or down? Broadly speaking, I would say we're experiencing much less crime now than we did 20 or 30 years ago. Many crime types have been declining at a fairly steady rate since the mid 1990s and in more recent years, probably since we started to emerge from the pandemic, crime levels have broadly flattened out at that lower level.

But some types of crime are rising now, some are still falling, and some are changing in ways that reflect shifts in society, shifts in technology and shifts in policing.

MF

That's the complex and highly nuanced picture, and it's the one that is designed to best serve those who make policies around crime, those who try to contain crime, those who try to fight it?

NS

That's right, Miles. And it's a picture that we get from drawing on several different data sources. There are two main ones. One is police recorded crime, and the second one is our independent survey of crime across England and Wales. And then we can use other data sources to provide richness for certain crime types, or to triangulate what we're seeing in those main data sources. And when we pull all of that together, we try and give a rich, nuanced, accurate picture for policy and policing.

MF

That's the aim of the statistics, but when it comes to public debate and public perceptions, do we risk misleading people by not being able to come up with a single barometer of crime? You can't go on the ONS website and see whether overall offending is up or down for example, or is that a completely pointless exercise?

NS

Well as I said, different crime types will have different things acting on them at any one point in time. But what we can do, for example through the crime survey, that has measured what I'm going to call traditional types of crime experienced by us as the public. So that's things like theft, that's things like violence, that's things like criminal damage, and in the last 10 years or so, that also includes fraud. And when we look at those types of crimes, we can see that, if you want a single figure, the numbers have come down. And that's when I say that over the last 30 years, there's been a big reduction in crime. If you take violence, theft and criminal damage, about 30 years ago, four in 10 of us, about 40% of us, every year would experience one of those types of crime. Now it's one in 10 of us. So, I can give you that picture for certain types of crime, but there are different ways of measuring and different data sources are better for certain types of crime, so coming up with that overall number is actually quite difficult.

MF

John, when it comes to political and media debate around crime, there are no simple answers, and yet those are to arenas where we want simple answers.

JOHN RENTOUL

Well, I think Nick did give a fairly simple answer, which is that if you ask people an open-ended question - have you been the victim of crime over the past 12 months - then the number of people saying yes to that question has gone down very dramatically over the last 30 years. So, in that very simple sense, crime has gone down hugely over the past 30 years.

But of course, people don't feel that, because that requires comparative memory...collective memory over a long period of time, and people are worried about what they read in the papers and what they see on social media. So, people, just as they always think that Britain is becoming a more unequal society, they always think that crime is rising, and it's very difficult to contradict that with simple statistics.

MF

Isn't that because there's always some aspect of crime, some type of crime that's always rising and it's opposition politicians, headline writers, particularly...of course you'll find lots of sophisticated, nuanced debate in the media...But those who write the headlines like to seize on the negative, don't they? Bad news sells. And you can see how people get these impressions because it's just the scary stuff they're hearing about.

JR

Well, possibly, although I think it's probably deeper than that. I think it's just human nature to feel fearful about the threats in society and the way of dealing with that psychologically is to assume that those threats are worse now than they used to be.

When you ask people has crime increased, they're not really giving you a statistical answer. They just say yes I'm afraid of various sorts of crime. And, you're quite right, the sorts of things that stick in people's minds are phone thefts and shoplifting, the sorts of things that get highlighted on social media all the time.

MF

As they used to say at the end of Crime Watch every week, these types of crimes are rare, don't have nightmares. Yet, that's no good if you've had your phone nicked or witnessed a shoplifting incident.

JR

That's right, and what's interesting about those two is that witnessing a shoplifting incident wouldn't be recorded in the crime survey because you personally are not the victim of that crime. So that's an example of an incident that has gone up, but that wouldn't be captured in the survey statistics, although having your phone nicked is something we probably would remember.

MF

And how responsible would you say our political leaders have been over the years? How responsible perhaps are they being now in the way that they present crime statistics?

JR

Well, it's very difficult, isn't it, because it's always partisan between government and opposition because ministers are always saying that crime has gone down, just as they're saying more nurses and doctors are employed in the NHS and all that. It's one of those statistics that well, certainly for the past 30 years, has been true, but opposition politicians have to try to argue the opposite, and they point to the sorts of crimes that have gone up, such as shoplifting and phone theft. So, it is a constant battle between what sounds to the public like just rival political claims, and the public will just discount more or less what any politician says and just choose to believe what they want to believe.

MF

Like many big statistical topics, once again, people can argue diametrically opposing things and both be right in a sense...

JR

Yes, exactly

MF

...and have some statistical basis for saying it. So it's the job of this podcast to help people untangle those sorts of complexities and decide for themselves.

Let's embark on a little journey then around how the crime statistics, these are the ONS crime statistics for England and Wales, how they are put together, how best to interpret them. And nobody better to guide us through that than yourself Nick, as joint head of crime. Starting at the beginning, because this would have been the original source for crime statistics going back a very long time indeed, and that is police recorded crime. And that sort of conjures up an image of The Bill, doesn't it, or Dixon of Dock Green for older listeners of a desk sergeant sitting there and dutifully recording offenses. Is that what it's like?

NS

I'm pretty old myself Miles, but that's stretching it...

MF

[Laughter] Yes. Very elderly...

NS

But yes, police recorded crime is one of the two main sources for crime statistics, and we report what we find in that data every single quarter. It's why John can confidently tell you that shoplifting, for example, is a current issue because we are seeing record levels of shoplifting offences in police recorded data quarter after quarter at the moment, so something's going on.

But we've got to remember that police recorded crime is dependent on a couple of things. The first thing it's dependent on is what we as the public report to the police. And the second key thing is that, even if we report that to the police, how do the police record that. If we go back 20 odd years, the national crime recording standard was introduced to police recorded crime. And this introduced concepts like if I report a crime, then I am to be believed, and that crime should be recorded. It shouldn't be the case that the police officer waits to find corroborating evidence, or thinks to themselves there's not much to go on there, I'm not likely to get very far with this and not record it. And there are some rules around that. So I might come and tell you that something's happened to me on a number of occasions, the same person has done something to me on a number of occasions over the past few weeks, that will be recorded as one crime. If I then a week later come and tell you it's happened again, that's a second one, but that first one that could have included lots of different instances, that's just one crime. Then there's a kind of weighting that goes on. So, when we count these things, there's something called the Home Office counting rules. And the most serious crime is what gets counted. You can have an incident that might involve violence, theft, criminal damage and, ultimately fraud, but it'll be the most serious of those crimes that gets counted.

Now what happened is, after the introduction of this standard at the start of this century, that really meant you couldn't use police data to measure trends in crime before that, because here was something that should start the current count again. Now, fortunately, we have the crime survey, which was an independent survey to go alongside it. But for the next 10 years, what we saw is that crimes recorded through police recorded crime dropped a lot faster than crimes that were recorded by the crime survey. And about 10 or 12 years ago, that led to people questioning police recorded crime, that led to some detailed audits of police recorded crime, and that led to conclusions that the police were actually not recording everything they should be.

And so what's happened in the last 10 years is real dramatic improvements to police recording crime, and we now get different data patterns. So for the last 10 years or so, police recorded crime numbers have gone up, but we understand why. It's because of improvements to police recording practices. At the same time, crime survey estimates have continued to trend down, and that's where we can come and use other data sources to triangulate against those two main data sources. So if I look at data from similar countries like Scotland and Northern Ireland, they match the crime survey data we have here. They don't match police recorded crime data. And if I look at Home Office outcomes data, which looks at the number of people that are actually charged or summoned for these offenses, the difference over the past 10 or 12 years matches crime survey data. It doesn't match the number of things that are recorded for police reported crime. And that's one of the key things that allows politicians who want to paint a different picture to seize on certain statistics that suit their agenda.

MF

Improved reporting was bound to lead to an apparent increase, wasn't it? So, it's that point that's got lost. It's because of the improved reporting, and it's been misinterpreted, John, just to bring you in on this, how widely understood is that point, or has it been wilfully ignored in some quarters?

JR

It's not widely understood, partly because there's such a strong belief among the public that crime is rising, that any evidence which appears to support that, such as police recorded crime, tends to get many more shares on social media than the crime survey graph showing that the level of crime is going down.

[Transition music]

MF

So there we are, a crucial and vital source of information there in police recorded crime. John, do you think there's anything the ONS should be doing to help people better understand the strengths and limitations of police recorded crime?

JR

I would say that what the ONS does on crime is a model of its kind actually. The reporting of that data is very carefully done, very soberly done, I mean, in a way, almost too restrained, because I think it sort of allows people to cherry pick the little bit that supports their argument rather than the bigger picture.

But no, I don't think there is anything much more that can be done, apart from trying to explain how the crime survey works in sort of simple language. But I mean the problem is that, generally speaking, people's understanding of probability surveys, representativeness and weighting of such surveys is not high. And if you say that the evidence that crime is much lower now than it used to be is from an opinion poll, then people will say, well I've never been asked, you can't trust them, and it's all done by You Gov and they're owned by the Tory party. I mean, it's just very difficult to explain to a lay audience how a proper representative sample survey works.

MF

Explaining the statistics and communicating them as well as possible, that'll always be an important priority for the ONS, but also, as we said earlier, making sure the policy makers, making sure the police, the experts, academic researchers and so on, have that detailed picture is half of the mission as well.

Let's turn then to the other big source of information about crime. It's one that we've discussed a fair bit already, but let's really unpack it, and that, of course, is the Crime Survey for England and Wales. The crime survey seeks to produce a snapshot of crime as experienced by the entire population, Nick?

NS

Yeah, that's right, and that removes some of those key variables that are in police reported crime. So, it's a national survey, we sample addresses around the country, we weight the data back to what the shape of the population looks like, and it's very in depth. So, it's not just anecdotal, it's rigorously designed, it's nationally representative, and it's been running for over 40 years now. Every year it involves tens of thousands of detailed interviews with members of the public, and the basic methodology has remained unchanged which is why it's so good for measuring trends over time.

And what we're asking people about in terms of their experiences of crime is that we don't care whether they've reported it to the police or not. It's what's actually happened to them, and if they tell us about it, we will record it, and we will assess whether it meets the threshold for a crime or not. So, it's that independence from the police data that's key. It's removing that influence of reporting behaviour and recording practices to try and give us a much clearer picture of actual victimization.

MF

And how representative is it at the moment, because much has been made elsewhere of the problem ONS has unfortunately been having getting people to fill in the Labor Force Survey, are response rates for the crime survey a better story?

NS

Well, response rates for the crime survey are one of the, if not the best in the country. But that isn't to say that we haven't seen similar impacts from the pandemic.

It's the pandemic that seems to have been this big rift that's changed things. So, for example, on the crime survey prior to the pandemic, about seven in ten addresses would eventually give you responses, and currently it's just under five in ten, so just under 50%. So, we have seen that big drop, but it is still a good response rate, generally speaking, and one of the best across the country. But what we have done, is we have checked the shape of the sample, in terms of completed responses, matches the last census, and it still does very closely. But we're keeping a close eye on it, because there could be things around crime that mean that the more people don't take part, the more chance there is for non-response bias, a technical term, to creep into results, particularly for certain types of crime.

MF

And that representativeness is so important, isn't it, because some groups are more likely, unfortunately, to be victims of crime than others.

NS

Absolutely right. Yep. So I mean, if you think about sensitive crimes like sexual offenses, you will see that younger people, particularly younger women, are more likely to be victims. If you look at things like violence or theft from a person, it's those types of people that tend to be out and about more, which again, is often younger people. And that's another example of where societal changes since the pandemic may have had an impact, and demographic changes over the past 10 or 20 years. So we've got an aging population. We've also got a population that perhaps doesn't go out as much as it used to. That reduces opportunities for crime and it reduces the demographic types of people that are more likely to both commit and be the victims of crime in the population. So there's lots of things going on underneath that we start to see reflected in the results.

MF

Does it tend to produce a less dynamic and less rapidly changing picture of crime?

NS

Yeah, it does, compared to police recorded crime. So, every reporting period that we report later from the crime survey will be based on, I mean, those interviews will have been asking respondents about their experiences in the year before that. So, in effect, each reporting period is covering about two years' worth of time, so there is more of a lag effect. That's one of the key things that police reported crime is good for. It's much quicker. You could tell that in the pandemic. Police recorded crime dropped very sharply, very quickly, and recovered relatively quickly afterwards, whereas crime survey data was much more of a slow pick up.

MF

And that John is, I guess, why the media and political commentators seem to be keener on the story being told by police recorded crime?

JR

It hadn't occurred to me, actually, that it was more sensitive to changes and would show changes more quickly. The media just responds to any dramatic negative change. I don't think we worry too much about the methodology behind them.

MF

Well, I mean, do people bother at all about the methodology behind it, because Nick has shared what a giant enterprise this really is. Is the value of that really understood in political debate?

JR

I'm afraid not.

MF

What can we do to underscore that?

JR

Yeah, I don't know. I mean, I think, I think it's just a matter of making the crime survey better known, because almost every debate I've taken part in on social media always goes along the lines of you can't believe the figures these days, nobody bothers to report crime anymore...And then, you know, I point out that it's not police recorded crime I'm talking about, that it's the crime survey. And then people say, oh well you can't believe government. Certainly, it is going to be always an uphill battle on something like crime, where public perceptions are very deeply entrenched, very passionately felt, and immune to facts of this kind. I mean, I think that's just something that you've got to live with and try to educate people as best you can.

MF

Ah, another thing the ONS is going to have to keep on at clearly. Okay then, talking about the facts, and this is the nub of the debate that's been going on recently in UK politics, about whether crime is up or not. So, give us the big picture. You've hinted at this already, but what is the big picture, in terms of what has changed in recent decades?

NS

Yeah. So the big change in recent decades is those traditional types of crime. So the traditional thefts, burglaries, vehicle thefts, used to be huge. Honestly, huge amounts of this stuff used to go on all the time, and it's things like that that have really dropped. So as I said, if you look at theft, criminal damage, and violence, with or without injury, from the crime survey, about 40% of us 30 years ago, that's four in 10 people, experienced at least one of those every year. Now it's about one in 10 or 10%. That is a massive, seismic drop. A generational drop. And that's from the crime survey. Now police reported crime showed exactly the same picture until 10 years ago, when there were those improvements, and for things like theft, it hasn't shown much of a jump since then, because they were relatively well recorded. You tended to need a crime reference number for your insurance claim for example. Whereas for things like violence, that's where we've seen a huge jump, and particularly for lower level violence. So it used to always be the case that the police would record a lot more violence that involved injury than they did violence that didn't involve injury. But in the last 10 years, that has completely flipped around. Now there's a hell of a lot more violence that doesn't involve the injury that gets recorded. What you're not telling me is it's the nature of violence in society that has changed, and we just don't hit people as hard anymore. And you know, of course, that's not what's happened. It's the way that things are being recorded that has changed. We also see a big increase in the number of stalking and harassment offences recorded from virtually nothing to 10 to 12 years ago.

MF

Is that because of greater awareness? That people are more ready to report these things?

NS

That's right. And there's also legislative change, changes to the law, which means that these things now are more likely to be considered offenses. And there was a period of time for a few years where the instruction to the police was don't only record stalking and harassment, but record the other crime that take place alongside it. So there was another wrinkle in the counts that came in, that has since been rescinded in the last couple of years, and we're starting to see it fall down again. One other reason why you can't look at trends. But what we do see in police recorded crime, what it's good for, is the most serious types of offenses and things that the crime survey just cannot pick up because we're not asking shopkeepers, for example. So shoplifting is the one.

So shop shoplifting is the one that we're at record levels for now, and have been every quarter for the past year, or 18 months or so. There's over half a million shoplifting offenses recorded every year now by the police. And you think that's quite a lot, but actually, if you were to look at the British Retail Consortium today, their estimate is that there's 20 million shoplifting offenses every year. Now, they don't publish their methodology. They do some sample of their members. I can't vouch for that figure, but let's say the number is somewhere in between the two, and let's say it's 10 million, because that's an easy number to work with. So if we're getting half a million offenses recorded by the police, but there's actually 10 million offenses. As shopkeepers, hopefully over time, start thinking actually, the police are taking this much more seriously, it's much more worthwhile me reporting this to the police, then you might actually see an increase in police recorded shoplifting offenses that is just an artifact of people being more likely to report it, rather than any change in the underlying level of crime. And similarly, that could happen the opposite way round as well.

So we do see that shoplifting is clearly up, but things like homicide, very flat, if anything, trending down over time, and things like theft from the perso. With phone theft that we've mentioned once or twice, we're seeing that spike in police recorded crime data, but we're also seeing it go up in crime survey data, particularly in London at the moment, through police recorded crime.

MF

Of course, one thing we haven't touched on much so far is the apparent rise in cyber-crime, very high-profile firms, brands, big name companies, getting hit. Secondly, the huge number of attempted frauds. I mean, just one example, today I get an email from a dodgy email address inviting me to renew my Spotify subscription. I haven't got a Spotify subscription, so clearly fraud there, but I'm not going to go report that to anybody, am I? Is it the case that, as some people might say, the villains have gone online in the last decade or so?

NS

Yeah, I think I would probably characterize some of the broader changes in crime over the last few years being that more of it, in relative terms, tends to take place either behind closed doors or online. And your description there of cyber-crime very much fits that bill. So about 10 years ago we developed a new module for the crime survey, which specifically asks people about fraud and computer misuse offenses, and it now makes up almost half, if not about half, of the individual crimes that we measure through the crime survey. They are some of the most common types of offenses people face, and we have adapted to try and include them. So for example, last year, the estimates are that nearly 1 in 12 of us experienced fraud or computer misuse where we were the intended victim. So the example that you describe where someone's asked you to confirm your Spotify account and you haven't got a Spotify account, at that point, as long as you don't click on something or go down their rabbit hole and into their dark world, you are not yet the specific intended victim. You're just one part of a big phishing exercise. It's if you click on that link and you end up being the specifically intended victim, you may or may not end up losing money or losing your card details. At that point, you start to count in terms of a victim of that type of offense. So it's difficult to measure. It does involve quite a lot of questioning. And the police measurement of fraud is patchy, I mean, the Office for Statistics Regulation did a review of police recorded crime on fraud and said there's a lot of room for improvement, basically. So the crime survey is a much better source for that data. But they're hard to detect, they're hard to report, they're hard to measure, and that is one area where we need continued investment in data quality, and we need continued, constant investment in public awareness, because those types of crimes and the tactics that criminals are using are changing all the time, and I wouldn't be surprised if this is one of the factors that has led to things like drops in response rates to surveys. I'm guilty of it. I will not answer the phone if I don't recognize the number. So it is leading to us being much more susceptible to being cynical.

MF

That's an interesting point. John I know you've taken part in an often charged social media debate around this where you've advanced the broad narrative that is told by the statistics, and yet so many people ready to doubt that. Is that because people do feel that crime is much closer these days? Now that we are getting the dodgy calls it feels much more proximate?

JR

I'm not convinced about that actually. I think the reason that cyber-crime is as a separate module on the crime survey is because it wasn't traditionally thought of as crime. I mean the sort of crime that people worry about is the violence and lawlessness on the streets, being attacked or burgled. I think it's very much to do with real world crime rather than computer crime.

MF

Or perhaps being more likely to witness shoplifting at first hand?

JR

Shoplifting and phone theft are the two things that really worry people and make them think that there's something to this idea that Britain is descending into lawless mayhem, but actually, we're safer and better off than we've ever been before.

MF

So Nick, what comes next in terms of how the ONS measures crime? Is it a question of refining these excellent data sources, and finding more corroboration just to improve their accuracy? Or are there game changing developments that might be available through technology or any other means that might not just improve the measuring of crime, but deal with some of these communications issues, and these trust issues, as well?

NS

Yeah. I mean, there's a few things there. So one I mentioned earlier is that we always need and try to keep up to date with the questions we're asking, the way that we're managing and tracking things. So as technology starts to have an impact on the types of crimes that people experience, we update and we adapt the questions. As the law itself changes, we update and we adapt the questions, and we'll continue to look at that.

Just picking up on your last point around perceptions, and why they feel perhaps differently to what the stats show, we ask that as well through the crime survey. We asked people whether they perceive crime to be going up, and we ask them whether they perceive crimes to be going up in their local area and at the national level. And there's this persistent gap, and it's quite a big gap, between what people think locally and what people think nationally, which you know either points to things like, it's the kind of things that you see on the street, or it's the kind of things you're seeing on the news, it's the kind of things you're seeing on your social media feed that are sort of giving you that emotional reaction that you think, well, nationally speaking, crime must be going up, even though I might not be seeing so much of it in my local area. So we will continue to evolve those things. The government of the day are particularly interested in looking at things like violence against women and girls, neighbourhood policing and better policing of crimes that are currently showing upticks, crimes like shoplifting and theft from the person. So there is some demand for more granular data, at sort of police force area level, a more local level than we can currently provide through the crime survey. So what is next is seeing whether technology can help us in a cost-effective way, to interview more people in an online environment. And that requires a lot of testing, and there's no guarantee of success, but that's one key area that we're going to be looking at. What's the art of the possible over the course of the next sort of couple of years or so.

MF

John, what would you be your advice to, well, firstly, people who want to understand that fully nuanced picture of crime, and listeners of this podcast, of course, are relentless seekers of that, but also to ONS on how to provide it for them, and perhaps how to try and allay people's irrational fears and better inform the topic.

JR

It's difficult, isn't it, because I think we are dealing with some quite powerful social forces of irrationalism and belief about the nature of society. And as Nick said, people think the country is going to the dogs even if their own local area is nice, secure, safe, quiet. I mean, there are some remarkable figures, which I think are from the crime survey, that people feel much safer walking around after dark in their local area than they used to, but people simply will not believe those kinds of data. So I think there's quite a lot of thinking to be done. I mean, on my part just as much as anybody else's as to how to convey a true and honest picture of what's actually happening to an audience which is just psychologically resistant to wanting to hear it. And I think that's got to do with acknowledging people's fears and saying, yes, shoplifting and phone snatching is going up, although, I mean, phone snatching is an interesting one. I don't think, I mean, I stand to be corrected on this, but I think, actually, it's not as bad now as it has been in the past. And there was a time, maybe 10 years ago, when phone thefts in London were very bad, and by working with the phone companies to make phones, in effect, disabled the moment they were stolen, that was brought under control. There's this sort of constant technological battle between the criminals and the phone manufactures. But you've got to acknowledge that people think correctly that some kind of crimes are increasing, although the ones that people are most worried about, such as knife crime and so on, I think the evidence is that it's not. But you've got to find some way of acknowledging people's fears before you try to get the actual information across.

MF

So, I hope we've made the point that the reality of crime is far more complex than a single headline can possibly convey. That's why understanding the full context really matters and we hope this podcast has been helpful to that end. Thank you to our guests and thanks as always to you at home for listening. You can subscribe to future episodes of Statistically Speaking on Apple podcasts, Spotify and all other major podcast platforms. I'm Miles Fletcher and from me and our producer Alisha Arthur, goodbye and mind how you go!

An inside view of the AI boom, with Read AI's David Shim

GeekWire

Agents hype versus practical, targeted automation

The AI-powered Podcast Player