How Ohio's train disaster could change the rail industry

Transcript

NIALA: Good morning! Welcome to Axios Today!

It’s Thursday, February 23rd.

I’m Niala Boodhoo.

Here’s what we’re covering today: AI chatbots and the risk of misinformation. Plus, a coast-to-coast winter storm. But first, how the Ohio train disaster could change the rail industry. That’s today’s One Big Thing.

How Ohio's train disaster could change the rail industry

NIALA: Transportation Secretary Pete Buttigieg is scheduled to visit East Palestine, Ohio today. A train carrying hazardous chemicals derailed there earlier this month releasing toxins into the air and water.

PETE BUTTIGIEG: I gotta tell you, ever since I came into this job, I have seen the power that multi-billion dollar railroad companies wield and they fight safety regulations, tooth and nail. That's gotta change.

NIALA: That’s Buttigieg speaking on Good Morning America this week calling on the rail industry to make changes in order to prevent these types of disasters in the future.

Washington Post’s Transportation Reporter Ian Duncan has the details on these reforms. Hi Ian, welcome to Axios Today.

IAN DUNCAN: Hi, thanks for having me.

NIALA: Ian. First, do we know more about what led to this accident happening to begin with at this point?

IAN: We don't know a ton. The NTSB, the National Transportation Safety Board is leading the investigation and they're pretty methodical. They've put out some information suggesting a wheel bearing on this train overheated. They have some video evidence that suggest that, and today they're going to release their preliminary report, which should have some new facts about the derailment and possibly the initial response as well. That won't sort of definitively say, here's why this happened, but it should give us some good clues about what went wrong.

NIALA: Specific policies in the industry have been blamed for this accident. What do we need to know about that?

IAN: There's a general sense that the industry has shifted to this model known as precision scheduled railroading, which is sort of an efficiency drive. So the industry used to rely really heavily on big trains that would carry all one type of stuff. And now what you see, more commonly, are trains like these ones that derailed in East Palestine that is carrying all kinds of different stuff and they're using new technology to try and move individual cars of goods where they need to be as quickly as they can. So it means cutting staff, and using longer trains potentially and if you talk to union people in the rail industry, they worry that that is just putting the industry under stress generally. The idea is that it's much more efficient. The, the criticism comes that the efficiency has then potentially led to kind of corner cutting when it comes to safety as well.

NIALA: One thing we've talked a lot about on Axios Today is just the lack of paid sick time for railroad workers. I know that Buttigieg announced some reforms to railroad safety earlier this week. How did paid leave factor into that?

IAN: That was one of the things that he called on the railroads to do immediately. His view is that if you have workers who can take the time off when they're sick and get better, they're gonna be safer when they're on the job. If you remember, this was a big point of discussion, leading into a potential rail strike in December, and the administration wasn't able to guarantee this leave then. But in the last couple of weeks, some of the railroads have started striking deals separately with their unions to provide that so CSX and Union Pacific, two of the big railroads have done that. So there has been some movement there.

NIALA: President Biden also tweeted that Congress should be involved. How so?

IAN: One of the big issues has been around this technology called electronically controlled pneumatic braking. And the idea there is that the existing brakes send a signal that travels down the train and so the cars can't all stop simultaneously. This technology is designed to allow to do that. Congress in 2015 told the Obama administration to take another look at whether that technology was justified. And the Trump administration ended up rolling back a rule that would've required it at least on some trains. So the administration now is saying maybe Congress can step in and help ease the way forward with that. The other thing that they're looking for Congress to do is to raise the cap on fines. It's just about $200,000 fine that railroad can face, and Northfolk Southern made $12 billion or more in revenue last year. And, and the administration's view is that those fines just are too small to be deterrent to a company of that size.

NIALA: I wonder if you can put the East Palestine derailment in context for us. How much of a catalyst do you think this may be for change in the railroad industry?

IAN: It's clearly generated a, a huge amount of attention. Secretary Buttigieg seems to think that that is gonna potentially open the door for some action in Congress. But that this is a huge industry that has a lot of inertia behind it and that has been putting its weight against regulation. They pay in high lobbyists, like lots of industries. Under the Trump administration, they were basically invited to come forward with wishlists of rules that they wanted while back. And they were successful in getting some of that. And so, they're gonna stick by those positions potentially.

NIALA: Ian Duncan covers transportation for the Washington Post. Thanks, Ian.

IAN: Thanks so much.

A coast-to-coast winter storm

NIALA: And in other transportation news: an enormous coast to coast storm has forced airlines to cancel thousands of flights — and more than 75 million Americans were under winter storm warnings yesterday.

The National Weather Service says the storm will ultimately affect every region of the lower 48 states - with either winter weather, severe thunderstorms or extreme temperatures likely through at least tonight.

Stay safe everyone and as always, this is a good reminder to check on your neighbors or others who might need some extra help in severe weather.

In a moment: what chatbots could mean for online misinformation.

[AD SPOT]

AI chatbots and the risk of misinformation

NIALA: Welcome back to Axios Today. I’m Niala Boodhoo.

AI tools like OpenAI's ChatGPT and Google's Bard have been the talk of the tech world for their remarkable abilities…BUT they could also spread dangerous misinformation far and wide.

Axios’ Sara Fischer reports generative AI doesn’t have a clear sense of the boundary between fact and fiction – and experts are sounding the alarm. Hey Sara.

SARA FISCHER: Hey Niala.

NIALA: Sara, I think we've all seen some fairly unsettling examples of chatbots like Microsoft's chatbot professing its love to a New York Times reporter. What are experts saying about AI bots and their potential risks?

SARA: Niala, this reminds me a lot of social media a few years ago where the biggest misinformation threat wasn't everyday users taking things out of context and spreading misinformation unintentionally. Although that is a huge threat. The biggest threat right now is that because it's highly unregulated and it's a new technology, bad actors are very easily able to exploit it. And so what would that look like? It might mean that a bad actor, maybe like a foreign adversary, uses misinformation to flow into AI models so that the answers that are popping out on the other are not accurate.

Also, we are looking at things like injection attacks where these bad actors or malicious users can teach lies to the programs, to the actual algorithms, which then can spread misinformation. Now, that's pretty highly technical, but it's a serious threat when you're looking at an upcoming election in a year and a half. Bad actors who have a history of using technology tools to spread misinformation, it's a huge and pressing problem.

NIALA: And on top of all of that, Google had an embarrassing moment recently when their Bard bot got a historical fact wrong in their public marketing video meant to show off the tool. What are tech companies doing to try to prevent this, especially misinformation?

SARA: Oh my gosh. It was a huge problem, by the way, with Google, Niala, because it shaved a hundred billion off their market cap in a day. I mean, these problems have huge business implications, and that's why tech firms are sort of sprinting to get ahead of any regulatory pressure or industry concerns around AI. One of the things that OpenAI did, which is the company behind Chat GPT, is it actually released a free tool that anyone can use. That's designed to help people understand whether or not a chunk of text was actually created through Chat GPT, or if it was created by a human. So let's say you're an educator and you're grading an essay, at least you have a tool to be able to see whether or not this was actually created by the student.

You also have tech companies like Google that are starting to issue guidance around their search results, saying, look, we're going to extra caution, and you should use extra caution when you're looking at search results around important topics like healthcare or financial information. Because in the new world of AI, we need to have extra, extra caution around what we're trusting.

NIALA: Sara, we know that regulation often lags behind technology, but is there anything in the works to help combat misinformation when it comes to chatbots?

SARA: Not in terms of actual regulation from our government. You know, one of the experts I spoke to said I've been having serious conversations with folks trying to get outreach to government and I'm having a hard time getting through. I think you're gonna continue to see innovation without regulation, and it's gonna be both the most beautiful and chaotic period in our tech history.

NIALA: Sara Fischer is Axios’ media reporter. Thanks Sara.

SARA: Thank you, Niala.

NIALA: I’m Niala Boodhoo - that’s it for us today! Thanks for listening, stay safe, and we’ll see you back here tomorrow morning.