Charlotte Gartenberg (3:32)

So we're not making the connection that rallying stocks is going to mean we'll soon really see EV tolls take to the air.

John Cindreou (3:39)

It's possible. Joby, for example, they've been at it for many, many years, far before they spacked. And they've done more than 100 test flights with a pilot. And these companies are starting to get results and get closer to maybe something that could be feasibly certified. But one has to remember that it's very complex. We're talking about aircraft that have to fly no matter what the weather is, aircraft that have to keep some battery life in reserve in case something goes wrong. Right now, what the FAA said is that they'll have to keep 20 minutes of flight time in reserve, which is better than they thought they were getting. They thought they were getting 30. So it's possible, but else we are seeing a lot of failure.

Charlotte Gartenberg (4:20)

Let's talk about that. There are some other players in the field right now. What's going on with them and what are the roadblocks that they're facing?

John Cindreou (4:28)

When we brother the lens, we see a lot more failures, including with evetol companies, particularly in Germany. We had two big ivtol companies, Lilium and volocopter. Especially in the case of Lilium, there is definitely a big technological component to why it's failed. The company sent an email just a few days ago saying that it's almost certain that it won't be rescued from insolvency. One has to remember that battery technology is still in its infancy and it's progressing very linearly. So it takes many years, right, for the range to increase. And also if you want to fly one of these vehicles, you have to remember going from vertical to horizontal is very complicated, and it takes a lot of energy. There's also a sentiment element to this green flying used to be cool, now it's out of fashion and you see it with Airbus canceling their own evetol project. For example, Boeing absorbed a company they were backing called Wisk, and it's still to be seen what will happen to it. So it really feels like the whole zeitgeist of the industry is changing.

Charlotte Gartenberg (5:31)

That was heard on the Street Columnist John Sindhriu Coming up, the counselor is not in, but the AI powered well being companion is Can a hybrid chatbot help with student mental health in schools? That's after the break.

Drew Barvier (5:55)

Your data is like gold to hackers. They're selling your passwords, bank details and private messages. McAfee helps stop them. Secure VPN keeps your online activity private. AI powered text scam detector spots phishing attempts instantly. And with award winning antivirus you get top tier hacker protection. Plus you'll get up to $2 million in identity theft coverage, all for just $39.99 for your first year. Visit McAfee.com, cancel anytime terms apply.

Charlotte Gartenberg (6:31)

Sonar Mental Health the developer of an AI powered well being companion named Sunny is rolling out its hybrid model to school districts which are struggling to meet student demand for mental health services. Our family and tech columnist Julie Jargan is here to tell us more. Julie, how does this chatbot work?

Julie Jargan (6:51)

So unlike most chatbots, which are AI, only this one is a little bit different. It's part AI, part human. As much as the AI learns from human interactions, the humans that are in the mix are also learning from the AI how best to communicate with teens. There are people with backgrounds in psychology, social work and crisis line support who monitor all of the chats with students. And they can insert themselves and edit the chats or write their own responses if they want to. And the AI learning from prior interactions with teens has discovered ways that seem to resonate most with teens about communication. For example, they've learned that teens don't really like smiley face emojis. They prefer things like the melting face emoji. And they've lear talking to teens across the country. What types of local vernacular resonate most with teens in certain geographic areas? So the humans have learned from that how to tailor their responses to the teenagers that they're communicating with and who developed it. So it was developed by Drew Barvier, an alumni of Stanford Graduate School of Business while he was at Stanford. He and another fellow classmate at Stanford developed this. They wanted to provide something that would fill in the gaps in mental health care which are considerable when it comes to young people especially.

Charlotte Gartenberg (8:11)

How was the chatbot built?

Julie Jargan (8:13)

They built it on several different large language models, and then it was trained in certain psychological techniques like cognitive behavioral therapy by a team of clinicians who have background in mental health services so that it can ask the right types of questions to teens and understand when a teen might be in crisis.

Charlotte Gartenberg (8:32)

So kids are already interacting with Sunny. Now, how widespread is the use of this hybrid chatbot?

Julie Jargan (8:38)

So it's currently in use in nine school districts across the country, and there are 4,500 students that have this available to them in those nine school districts. And a lot of these districts are in low income and rural areas where there's a shortage of mental health services in general. And then of course, at the school level as well, with a shortage of school counselors. This is supposed to be something that doesn't replace humans, but when there aren't enough people for students to talk to, that this is something that could catch issues early on. And then, you know, if students are in need of mental health support as identified by the AI and the humans that are a part of this, or they're talking about self harm, that they can flag that to the correct people and intervene and get help before there's a crisis.

Ryan Reynolds (9:22)

So.

Charlotte Gartenberg (9:23)

And our CEO Drew Barvier told you that he makes it clear to schools and students that Sunny isn't a therapist and the bot frequently encourages kids to talk to the humans in their lives. So, Julie, what are some of the potential risks?

Julie Jargan (9:37)

There's always privacy risks with things like this, especially when you've got young people who are providing what could be very sensitive personal information. Sonar says that in its privacy policy that once a student closes out their account, they retain the data for 60 days and then delete it. And that parents and students can request to have any chats deleted at any time. But data privacy is always something that's a concern. In a situation where there could be hacking or something like that, having personal information out there is always a risk.

Charlotte Gartenberg (10:10)

So how is Sunny going? What are users saying about it?

Julie Jargan (10:13)

From the people I talked to, they said that they've gotten positive feedback from students that students feel like this is a sort of a judgment free zone where they can share their obsessive thoughts that might annoy their friends, that they can talk about the same problems that they're ruminating on over and over again. And Sunny is not going to tell them to stop, that it's always there to listen and steer them kind of in the right direction. Some of the schools I talked to said they've seen a drop in student disciplinary problems since they began using this. Others have said they've seen an increase in grades. It's still early days to get the results, and whether it's related to this chatbot or something else, but there's been generally positive feedback from the students who've used it.

Charlotte Gartenberg (10:55)

What kind of concerns do people have.

Julie Jargan (10:57)

About the bot when you talk to therapists? There's a concern that even though this isn't meant to be a replacement for therapy, that some students who might benefit from therapy just won't go that route because they feel like this is it, I've got someone to talk to and they won't take that extra step to have a face to face interaction. Therapists say that can an AI really understand if a student is experiencing depression, anxiety or other mental health issues? Unlike in a face to face interaction where a therapist can gauge body language and kind of take some of those nonverbal cues to understand how a student might be struggling and that could be missed by AI.

Charlotte Gartenberg (11:37)

That was our family and tech columnist Julie Jargan. And that's it for Tech News Briefing. Today's show was produced by Jess Jupiter with supervising producer Kathryn Millsap. I'm Charlotte Gartenberg for the Wall Street Journal. We'll be back this afternoon with TNB Tech Minute. Thanks for listening.