WEBVTT - ICYMI: Axon Sees an A-I Future 

0:00:02.520 --> 0:00:10.760
<v Speaker 1>Bloomberg Audio Studios, Podcasts, radio news. This is Bloomberg business Week,

0:00:10.880 --> 0:00:14.720
<v Speaker 1>Daily reporting from the magazine that helps global leaders stay

0:00:14.760 --> 0:00:18.800
<v Speaker 1>ahead with insight on the people, companies, and trends shaping

0:00:18.840 --> 0:00:23.760
<v Speaker 1>today's complex economy. Plus global business finance and tech news

0:00:23.800 --> 0:00:27.800
<v Speaker 1>as it happens. The Bloomberg Business Week Daily Podcast with

0:00:27.960 --> 0:00:32.080
<v Speaker 1>Carol Masser and Tim Steneveek on Bloomberg Radio.

0:00:32.240 --> 0:00:36.680
<v Speaker 2>Matt Miller here in the Bloomberg Are we in the

0:00:36.680 --> 0:00:38.880
<v Speaker 2>Bloomberg Business No, the Bloomberg Business Week Studio.

0:00:39.040 --> 0:00:42.159
<v Speaker 3>Sometimes I get lost, get lost. I don't know.

0:00:42.240 --> 0:00:44.680
<v Speaker 2>It's a big building, right we have twenty nine floors here.

0:00:44.800 --> 0:00:47.680
<v Speaker 2>I'm here with Emily Graffeo. We're in for Tim and Carol,

0:00:48.240 --> 0:00:50.640
<v Speaker 2>and we have with us right now Josh Isner, president

0:00:50.640 --> 0:00:54.800
<v Speaker 2>of Axon, on the latest enhancements in public safety technology.

0:00:54.880 --> 0:00:58.880
<v Speaker 2>They make Taser guns, they make body cameras, but they

0:00:58.920 --> 0:01:02.280
<v Speaker 2>make a lot more than that, going beyond hardware into

0:01:02.320 --> 0:01:04.040
<v Speaker 2>software as well. Josh, great to have you on the

0:01:04.040 --> 0:01:07.760
<v Speaker 2>program with us. Talk to us first about the kind

0:01:07.760 --> 0:01:10.880
<v Speaker 2>of non lethal tools that the police and I guess

0:01:10.959 --> 0:01:16.720
<v Speaker 2>National Guard have to combat rioters, if you want to

0:01:16.720 --> 0:01:18.840
<v Speaker 2>see them that way. We've heard a lot of talk

0:01:18.920 --> 0:01:22.360
<v Speaker 2>out of the White House about an insurrection in Los Angeles,

0:01:22.440 --> 0:01:28.280
<v Speaker 2>and obviously nobody wants fatalities. So what kind of equipment

0:01:28.280 --> 0:01:31.640
<v Speaker 2>are you providing to these law enforcement officers to deal

0:01:31.840 --> 0:01:33.520
<v Speaker 2>with unrest?

0:01:34.640 --> 0:01:38.840
<v Speaker 4>Yeah, certainly, and we'll probably at least I will do

0:01:38.920 --> 0:01:41.040
<v Speaker 4>my best to kind of stay out of the politics

0:01:41.080 --> 0:01:44.559
<v Speaker 4>of the situation, focus on how we can keep people safe.

0:01:44.600 --> 0:01:47.720
<v Speaker 4>And ultimately that's what the taser does. It's a less

0:01:47.800 --> 0:01:50.960
<v Speaker 4>lethal tool, shoots out darts. Two of the darts connect

0:01:50.960 --> 0:01:54.800
<v Speaker 4>to your body and conduct energy at the same waves

0:01:54.800 --> 0:01:57.600
<v Speaker 4>that your brain sends to your muscles, and that's what

0:01:57.720 --> 0:02:01.160
<v Speaker 4>causes your muscles to tense up or five seconds it's

0:02:01.200 --> 0:02:03.680
<v Speaker 4>over and the police take you into custody. And so

0:02:04.600 --> 0:02:08.799
<v Speaker 4>you know, in Greater LA, all of those departments, we're

0:02:08.840 --> 0:02:11.400
<v Speaker 4>proud to be their partner in deploying less legal tools

0:02:11.440 --> 0:02:15.000
<v Speaker 4>to them and and they have those the tools, you know,

0:02:15.440 --> 0:02:18.640
<v Speaker 4>to hopefully safely de escalate some of what's going on.

0:02:19.080 --> 0:02:23.160
<v Speaker 5>Josh, have you seen demand pick up? It's very short term,

0:02:23.160 --> 0:02:25.480
<v Speaker 5>but just in the last several days, have you seen

0:02:25.840 --> 0:02:27.680
<v Speaker 5>that demand at least coming from.

0:02:27.600 --> 0:02:32.919
<v Speaker 4>LA Not not particularly only because the demand is already

0:02:32.919 --> 0:02:36.400
<v Speaker 4>there from all of those customers, they certainly have seen

0:02:37.720 --> 0:02:43.000
<v Speaker 4>riots like this historically, they've seen other, you know, applications

0:02:43.000 --> 0:02:47.120
<v Speaker 4>for these. Generally, when you have any kind of physical

0:02:47.120 --> 0:02:49.480
<v Speaker 4>threat to the officer or someone holding a knife, or

0:02:49.800 --> 0:02:53.359
<v Speaker 4>or sometimes even something you know more deadly than that,

0:02:54.000 --> 0:02:57.720
<v Speaker 4>there's there's very clear, you know, established use cases for

0:02:57.720 --> 0:03:00.480
<v Speaker 4>for using a taser. So really proud of the fact

0:03:00.520 --> 0:03:03.040
<v Speaker 4>that it didn't take something like this for our customers

0:03:03.120 --> 0:03:05.799
<v Speaker 4>to already trust us and believe that there are use

0:03:05.840 --> 0:03:07.040
<v Speaker 4>cases for our products.

0:03:07.600 --> 0:03:11.160
<v Speaker 3>I will say that I have tased myself on a

0:03:11.160 --> 0:03:14.120
<v Speaker 3>few occages. I have to say that.

0:03:14.200 --> 0:03:17.040
<v Speaker 2>So years ago I was in a job, like a

0:03:17.080 --> 0:03:21.400
<v Speaker 2>summer job, doing security and we had to use these tools,

0:03:21.440 --> 0:03:23.440
<v Speaker 2>and our boss said, I want you to try it

0:03:23.480 --> 0:03:26.240
<v Speaker 2>on yourself before you use it on anyone else.

0:03:26.280 --> 0:03:28.399
<v Speaker 3>And I never actually ended up using it on anyone else,

0:03:28.440 --> 0:03:31.760
<v Speaker 3>but I'm still here to talk about it. So, Josh,

0:03:32.040 --> 0:03:32.639
<v Speaker 3>I wonder.

0:03:32.440 --> 0:03:36.600
<v Speaker 2>About the importance of AI in the tools that you provide.

0:03:36.600 --> 0:03:40.240
<v Speaker 2>Emily and I were talking earlier about and I don't

0:03:40.240 --> 0:03:42.200
<v Speaker 2>know that we're there yet, but if you could use

0:03:42.320 --> 0:03:45.400
<v Speaker 2>AI and facial recognition, for example, on a body cam

0:03:45.600 --> 0:03:49.120
<v Speaker 2>or in smart glasses, then you'd give law enforcement officers

0:03:49.160 --> 0:03:50.680
<v Speaker 2>maybe a little bit of an advantage.

0:03:51.720 --> 0:03:53.400
<v Speaker 3>Have you thought about, I mean, I'm sure you've thought

0:03:53.400 --> 0:03:55.160
<v Speaker 3>about have you started to develop tools like that?

0:03:56.400 --> 0:03:59.160
<v Speaker 4>Yeah, so right now we do have a lot of

0:03:59.360 --> 0:04:03.600
<v Speaker 4>AI functionality built into our products and our cameras. I

0:04:03.600 --> 0:04:06.920
<v Speaker 4>think the tool we've focused on first with the body

0:04:06.960 --> 0:04:10.840
<v Speaker 4>camera is actually offloading all the audio transcripts of the

0:04:10.960 --> 0:04:14.840
<v Speaker 4>videos that take place and letting AI analyze those in

0:04:14.920 --> 0:04:18.440
<v Speaker 4>suggest a first draft of the police report to the officer.

0:04:19.480 --> 0:04:22.719
<v Speaker 4>There are some sexier use cases of AI, like the

0:04:22.760 --> 0:04:26.000
<v Speaker 4>facial recognition element you mentioned, but the use case I

0:04:26.120 --> 0:04:31.320
<v Speaker 4>just mentioned literally takes the officer's administrative workload down from

0:04:31.360 --> 0:04:35.200
<v Speaker 4>fifty percent to twenty percent, and thus an officer ends

0:04:35.279 --> 0:04:38.159
<v Speaker 4>up in the street with far more of their time

0:04:38.400 --> 0:04:41.320
<v Speaker 4>days of a week that we're giving back to the officer.

0:04:41.440 --> 0:04:45.920
<v Speaker 4>And so our early AI tools are much more focused

0:04:45.960 --> 0:04:49.760
<v Speaker 4>on productivity such that officers can do what they do

0:04:49.880 --> 0:04:53.919
<v Speaker 4>best at more often. But I think as we build

0:04:53.960 --> 0:04:58.279
<v Speaker 4>out our next generation body cameras with more frankly battery power,

0:04:58.440 --> 0:05:03.640
<v Speaker 4>computational power, better connectivity, you'll see some of those other

0:05:03.720 --> 0:05:06.880
<v Speaker 4>AI products at the edge start to start there to

0:05:06.920 --> 0:05:09.359
<v Speaker 4>make their way into the body cameras. I think first

0:05:09.360 --> 0:05:13.280
<v Speaker 4>it'll probably be license plate recognition. But over the longer term,

0:05:13.440 --> 0:05:15.840
<v Speaker 4>you know, once we can prove that these speci facial

0:05:15.880 --> 0:05:19.960
<v Speaker 4>recognition models are rid of inherent bias, that's when I

0:05:20.000 --> 0:05:23.400
<v Speaker 4>think we end up, you know, launching a product like that.

0:05:23.520 --> 0:05:27.360
<v Speaker 2>You know, Josh, the the ac LU has already brought

0:05:27.440 --> 0:05:32.960
<v Speaker 2>up concerns about the bias inherent in AI, the hallucinations

0:05:33.000 --> 0:05:39.520
<v Speaker 2>that AI is capable of, and I think has issued

0:05:39.560 --> 0:05:44.920
<v Speaker 2>complaints about use of AI and your software to draft

0:05:45.000 --> 0:05:47.599
<v Speaker 2>police reports. How do you respond to that, because of

0:05:47.600 --> 0:05:52.160
<v Speaker 2>course it can make mistakes and if you see a

0:05:52.240 --> 0:05:55.560
<v Speaker 2>first draft, even if you're a trained police officer, that

0:05:55.600 --> 0:05:59.800
<v Speaker 2>can also alter the way you think of what actually happened.

0:06:00.160 --> 0:06:03.440
<v Speaker 2>So is it maybe doing as much harm as it

0:06:03.480 --> 0:06:03.960
<v Speaker 2>does help?

0:06:05.400 --> 0:06:07.760
<v Speaker 4>No, I wouldn't put it that way. And and and look,

0:06:07.880 --> 0:06:09.560
<v Speaker 4>you know, before I get to kind of the process

0:06:09.600 --> 0:06:11.760
<v Speaker 4>and we run, I think it's reasonable to expect, like

0:06:11.800 --> 0:06:13.080
<v Speaker 4>the a c l U is never going to be

0:06:13.080 --> 0:06:16.240
<v Speaker 4>doing art reels about police technology or what you know,

0:06:16.279 --> 0:06:19.520
<v Speaker 4>what police you know, products they use and so forth,

0:06:19.560 --> 0:06:21.280
<v Speaker 4>And so I do think we have to remember a

0:06:21.320 --> 0:06:24.359
<v Speaker 4>little of that And and you know. Our approach to

0:06:24.520 --> 0:06:29.240
<v Speaker 4>this is is something called our ee a C, which

0:06:29.279 --> 0:06:32.920
<v Speaker 4>is our Ethic Equity and Ethics Advisory Coalition, and this

0:06:33.000 --> 0:06:36.040
<v Speaker 4>is a group internal to ax On of folks that

0:06:36.960 --> 0:06:40.800
<v Speaker 4>come from the world of activism and don't see the

0:06:40.839 --> 0:06:43.799
<v Speaker 4>issues the same way that the police do and see

0:06:43.800 --> 0:06:46.919
<v Speaker 4>issues more like the a C l U does and

0:06:46.920 --> 0:06:49.960
<v Speaker 4>and they're part of our product development process. So we

0:06:50.000 --> 0:06:53.680
<v Speaker 4>have gates along our product development where we check back

0:06:53.720 --> 0:06:56.480
<v Speaker 4>in and do a full product deep dive with our

0:06:56.520 --> 0:06:59.360
<v Speaker 4>ee a C and they ask those questions and they

0:06:59.400 --> 0:07:02.400
<v Speaker 4>want to see it demonstrated that these models are rid

0:07:02.440 --> 0:07:06.120
<v Speaker 4>of bias and so forth. And that's been a very

0:07:06.320 --> 0:07:12.640
<v Speaker 4>very productive relationship for us with our EAC, and it

0:07:12.960 --> 0:07:18.360
<v Speaker 4>really does validate internally, you know, our product designs and

0:07:18.440 --> 0:07:21.600
<v Speaker 4>motivations to make sure that hey, when these things hit

0:07:21.640 --> 0:07:23.960
<v Speaker 4>the market, you know, the folks who are the most

0:07:24.000 --> 0:07:27.800
<v Speaker 4>critical voices in the room will have the data to

0:07:27.880 --> 0:07:30.800
<v Speaker 4>share with them about the process we took, you know,

0:07:31.600 --> 0:07:34.360
<v Speaker 4>to arrive at these products. And you can read a

0:07:34.360 --> 0:07:37.080
<v Speaker 4>little more about this on Axon dot com. It's called

0:07:37.080 --> 0:07:41.560
<v Speaker 4>our Innovative Responsible Innovation Framework and it's on there and

0:07:41.600 --> 0:07:43.560
<v Speaker 4>you can see exactly how we think about some of

0:07:43.600 --> 0:07:44.200
<v Speaker 4>those issues.

0:07:44.480 --> 0:07:48.760
<v Speaker 5>So, Josh, your company started for primarily law enforcement. I'm

0:07:48.800 --> 0:07:52.920
<v Speaker 5>wondering if you're seeing more demand pick up from civilians,

0:07:52.960 --> 0:07:57.440
<v Speaker 5>maybe civilians worried about crime, particularly post pandemic. Have you

0:07:57.480 --> 0:08:00.760
<v Speaker 5>been seeing more civilians and what products clar are they

0:08:00.840 --> 0:08:01.320
<v Speaker 5>looking for.

0:08:02.160 --> 0:08:05.760
<v Speaker 4>We we do sell a consumer taser device and uh

0:08:06.160 --> 0:08:09.200
<v Speaker 4>and and it has picked up since the pandemic. For sure,

0:08:09.560 --> 0:08:12.080
<v Speaker 4>It's still a relatively small part of our business. But

0:08:12.160 --> 0:08:14.160
<v Speaker 4>over the next couple of years, you know, I think

0:08:14.200 --> 0:08:17.240
<v Speaker 4>we'll be updating our product portfolio in the consumer space,

0:08:17.280 --> 0:08:20.600
<v Speaker 4>and we're really excited about the future iterations of the product.

0:08:20.840 --> 0:08:24.320
<v Speaker 4>You know, Ultimately, our core market is police and we

0:08:24.360 --> 0:08:27.080
<v Speaker 4>wouldn't want it any other way. Where we're honored to

0:08:27.120 --> 0:08:29.600
<v Speaker 4>work with police, and we think our products just have

0:08:29.720 --> 0:08:33.200
<v Speaker 4>such great product market bid in that market, but in

0:08:33.280 --> 0:08:36.000
<v Speaker 4>consumers where folks don't want to carry a gun and

0:08:36.480 --> 0:08:40.120
<v Speaker 4>they're looking for you know, other alternatives that are that

0:08:40.200 --> 0:08:44.160
<v Speaker 4>can protect them but not in a lethal fashion. You know,

0:08:44.360 --> 0:08:47.000
<v Speaker 4>I think the taser is a viable alternative. So so

0:08:47.040 --> 0:08:49.720
<v Speaker 4>we're excited to talk more about that in the coming

0:08:49.800 --> 0:08:51.920
<v Speaker 4>years as we as we launched some new products in

0:08:51.920 --> 0:08:52.480
<v Speaker 4>that space.

0:08:53.360 --> 0:08:58.719
<v Speaker 3>This is the Bloomberg Business Week Daily podcast, available on Apple, Spotify,

0:08:58.840 --> 0:09:01.040
<v Speaker 3>and anywhere else you your podcasts.

0:09:01.440 --> 0:09:05.320
<v Speaker 2>Listen live weekday afternoons from two to five pm Eastern.

0:09:05.160 --> 0:09:08.880
<v Speaker 3>On Bloomberg dot com, the iHeartRadio app, tune In, and

0:09:09.120 --> 0:09:10.559
<v Speaker 3>the Bloomberg Business App.

0:09:10.800 --> 0:09:13.560
<v Speaker 4>You can also watch us live every weekday on YouTube

0:09:13.800 --> 0:09:15.960
<v Speaker 4>and always on the Bloomberg terminal