108 comments

[ 6.0 ms ] story [ 177 ms ] thread
Not gonna lie, pretty much the only reason I clicked the article was to see a video of rats driving tiny cars.
Thank you for your honesty.
I, too, prefer the silly video to the underlying scientific premise.
I watch the video and see the next generation of competitors who will be beating me at rocket league.
I was just about to complain about the missing video link. Thanks!
I just want to complain about them continuously helping the rat.

Backseat drivers.....

Thank you, that brought it to life much better than the article even though, of course, the article had a lot more background information.
Please link to a better video.
Scaling that up to rats driving your "self-driving" car, I don't think I'd feel safe with less than 5 rats and a consensus mechanism in-place.

Perhaps some of these rats could even be replaced with sensors and Machine Learning. Just imagine the possibilities

If the rats are able to communicate with each other then I think you might find the consensus would not necessarily be in your favour.

What happens if you get allocated five 'teenage' rats who just love pushing your Tesla to its limits! I mean it's scary enough when it's just me and the Autopilot on the autobahn at 240 km/h.

Shoo-in for the Ig Nobel
The story is not rats driving cars. This story is not that rats in enriched environments acquire driving skills more quickly.

The story is that learning to drive relaxes rats.

Furthermore, the rats that lived in enriched environments drove for the joy of driving, whereas the standard caged rats only drove for the food rewards.

"As hypothesized, the animals living in the enriched environment performed better at the driving test, indicating that they did a better job at learning a new complex skill. The enriched rats also maintained a strong interest in the car, even after the reward of food was removed.

On the other hand, the researchers were surprised at the lack of interest shown by the non-enriched rats and their level of underachievement shown in the driving task. "

Intrinsic vs extrinsic reward.

Oh, "interest in the car" meant interest in driving? Hard to know without pdf access.

"Just like us humans, learning to drive and navigate seemed to have a relaxing effect on the rats. In a control experiment, they found rats had higher levels of cortisol when being driven around in remote-controlled cars than when they were allowed to steer themselves."

From the futurism post https://futurism.com/the-byte/scientists-rats-drive-tiny-car...

Seems like a terrible control imo for me to say driving intrinsically has a relaxing effect. A better control would be rats doing nothing vs rats driving and seeing if the rats doing nothing were more stressed than the driving rats, meaning the driving rats were actually lowering their baseline stress.

In this experiment the control of being a passenger in a terrifying vehicle moving all by itself with no autonomy or control over the situation can quite likely be the thing causing elevated stress, rather than rats driving depressing levels of stress. Imagine if someone suddenly strapped you into a bubble that started moving by itself and you have no idea why this is happening, no control over the situation, or where you're going or what's going to happen to you. Stressful af. Hell, people get stressed just being in the passenger seat watching someone else drive.

The vast majority of people prefer having autonomy and control over their own motion vs being helplessly navigated by someone else you don't know/trust with zero context and no idea what's going on. A little misleading if this reflects the actual study.

"Rats are terrible back-seat drivers, scientists find."
A connection with Seligman's https://wikipedia.org/wiki/Learned_helplessness experiment

Two dogs in separate rooms with floors that can give them a mild shock. Arranged so they get exactly the same shock. One dog can turn off the shock by performing an action; the other has no control.

This experiment turns the second dog into a shivering nervous wreck... but the first is fine. Same shocks; only difference was control.

>A better control would be rats doing nothing vs rats driving and seeing if the rats doing nothing were more stressed than the driving rats, meaning the driving rats were actually lowering their baseline stress.

What if it's the opposite? What if rats that do nothing are under a higher stress and driving simply allows them to get back to their baseline stress?

That’s a different thing, the actual control was as you describe (measuring stress markers over time). The thing mentioned appear to be an attempt to confirm some previous study's finding about self-sufficiency, which is pretty much related to what you say about autonomy and control.

Never assume bad science when bad journalism would suffice.

And never assume bad journalism when you’re reading comments on HN where most of the commentators haven’t read the article they are commenting on.
I took my first driving lesson on an automatic today, and it was kind of a surreal out of body experience. I felt like I was playing a video game. Maybe because I could control something that was moving the world around me?
Until they put the rats in rush hour traffic, then see how relaxed they are :-)

This was a really a fun paper, I'd love to see it reproduced. Perhaps they could compare mice behavior to rat behavior, although given their size it might be more interesting to train a mouse to ride a tiny motorcycle.

Reminds me of the chapter of Blink where Gladwell delves into the correlation between a doctor's bedside manner / tone of voice and the likelihood of them being sued for malpractice. The whole thesis of the chapter seems to be that people are more willing to sue docs who are less nice, and then at the end he says if you don't like your doc, your intuition is probably right!
Wow! Let's train them a little bit more then put them in Formula-E race cars on a circuit :D
Forget that. I'd pay to see rat drone racing. Especially if it turned out that practiced rats flying local could often beat humans flying remote.
AMP-free URL that actually works without javascript:

https://www.iflscience.com/brain/teaching-rats-to-drive-tiny...

sigh.

The AMP link loads faster, the page is easier to read, and there aren't any ads.

I'll take the AMP link this time around!

The original site has a video of rats driving tiny cars, the whole reason I clicked in the first place. The AMP link does not.
Using noscript makes most sites faster and ad-free, all while getting rid of more untrusted code running in the browser.

I'll take the AMP-free web thank you very much.

Not really surprising. From what I have seen of rats, they are very good at spatial perception and physical manipulation of their environment, where they easily outperform most cats and dogs I have had the pleasure of knowing. A smart rat clearly enjoys exploration and challenge, and likes to be in control - it would appear more unexpected if these joyrides didn't relax them.
If this interests you, you may want to read Robert Sullivan’s Rats: Observations on the History & Habitat of the City's Most Unwanted Inhabitants.

It is all about rats and made me much more appreciative of them. Now when I see one I think about what it’s doing and why.

The book also has some interesting stuff about US history.

I remember a short-story where a cat learns to pilot a customized glider. It has always made me wonder about the plausibility of it. Were I not allergic to cat, I would probably have built a customized controller for a roomba and see if they are interested in learning to drive it.

I really wonder how far animals can in tool usage if we were to build custom-designed ones for them.

(comment deleted)
To make this work at all, you have to carefully tune your learning rat - @fchollet
The researchers should set up a company making the vehicles for sale. In various size perhaps for different animals.
"Teaching Rats To Drive Tiny Cars Helps Them Relax, Scientists Discover"

Yes, but the rats might be thoroughly stressed out

I have a driving licence, but I don't drive and haven't for years. It's way too stressful, being responsible for a high-speed machine and all the lifes of people passing by, not even speaking about my own or passenger's. I can relax in a train, but definitely not in a car ;p
There was no traffic or pedestrians in the videos.