Skip to main content
Johan Land

Johan Land

Oct 14, 2026

14:04
Which of those three is the hardest to get right?
14:07
I think the automation with agents is the hardest, and it's because analyzing the data is something that you still can do fairly deterministically and, and, and it doesn't require lots of, of, of judgment.
14:21
Whereas once you start deploying AI and agents into an operation, you need to start making decisions.
14:29
You need to make those decision based on data and based on analytics, but often there's a layer of judgment involved there, like a judgment call.
14:36
Examples of this would be...
14:38
Very serious examples would be you have an accident that has happened on, on, on the road.
14:44
What is the right next thing to do? Should you con- should, should you dial 911 on behalf of, of, of the driver because this is a serious situation? Or should you, should you contact, contact someone else, or should you check in with the driver through, through, through like voice? Or like what's the right next step, right? Like it's a, it's a very...

23 MINS LATER

38:11
[laughs]
16:49
Okay
16:49
...
16:49
but then OpenAI hears the rumor that there's a breakthrough, right? So a week ago, what OpenAI does [laughs] is that they start just, it looks like in desperation, spending, like, $10 billion plus on, like, 10,000 plus agents trying to solve the six remaining millennium problems, and they get traction on exactly Navier-Stokes, right? So then they...
17:11
So, but then [laughs] this gets even better, 'cause then, uh, Levent and Tristan, they hear the rumor, right, about OpenAI's had a breakthrough.
17:19
So then they reach out to OpenAI and, and, and here...
17:22
And all this so far is, like, fact.
17:24
It's well documented.
19:29
... help me understand the mi- the- the the mix- mismatch here.
11:21
So, um, and you're, and you're like your opinion, when did more visibility stop being the product?
11:27
Yeah, look, you're so right.
11:32
The way we think about this in the context of our customers, but I think it's true for the entire market.
11:37
We started with getting ourselves connected.
11:40
We connected data and started gathering it, uploading it to the cloud and whatnot.
11:44
I don't know, maybe that's 20 years ago or something.
11:46
It depends on different people in different stages.

8 MINS LATER

19:41
But if you were to kind of like strip away the PR, for how many of these jobs is that true for and like for how long?
10:59
How are you guys building that forward?
11:00
Absolutely.
11:01
Training is truly crucial.
11:02
All drivers want to do great.
11:04
The problem is that companies are struggling with having the bandwidth to train everyone.
11:10
So the way we see this going and what we have available already now and much of what we announced is AI helping augment people and do some of the training for people.
11:20
by some of the coaching.
15:22
Is there a project inside your business that you're most passionate about? Yeah, I mean...
23:35
What are fleets actually seeing real value out of when it comes to AI today?
23:40
Yeah.
23:41
And look, we have been deploying AI for more than 10 years, right? At scale, like at high scale.
23:47
We are on, you know, 20-plus percent of vehicles in the US, so, like, uh, in, in commercial fleets.
23:52
Like, it's, it's at truly scale.
23:54
And the types of things that we've been doing all along is effectively have the AI drive with people, observe what is happening, and then alerting drivers to it.
24:03
So that can be things like you're having a too close following distance.
28:31
How do, how does this impact those nucle- nuclear verdicts in risk management? You talked about the driver being exonerated, but from that discovery angle and that type of stuff, how, how is this impacted?
24:02
What are fleets actually seeing real value out of when it comes to AI today?
24:06
today?Yeah.
24:08
And look, we have been deploying AI for more than 10 years, right, at scale.
24:13
Like at high scale.
24:14
We are on, you know, 20-plus percent of vehicles in the US are like, uh, in, in, in commercial fleets.
24:19
Like it's, it's at truly scale.
24:21
And the types of things that we've been doing all along is effectively have the AI drive with people, observe what is happening, and then alerting drivers to it.
28:58
How do, how does this impact those nucle- nuclear verbs in risk management? You talked about the driver being exonerated, but from that discovery angle and that type of stuff, how, how is this impacted?

We value your privacy

We use cookies to understand how you use our platform and to improve your experience. Click “Accept All” to consent, or “Decline non-essential” to opt out of non-essential cookies. Read our Privacy Policy.