Electric motor? Just bought the car? That noise is the ac servo don't worry about it my car has done this for years. If you want to worry about something worry about the hybrid battery that feeds the electric motors. That's the most common thing failing on a G2. You have a very old G2.
There are two possibilities for a back-and-forthie noise like what you recorded there. One, as Ed suggests, is an HVAC servo. (That really counts as three possibilities, as the HVAC has one servo to move the air-inlet door, one to move the air-outlet selector, and one to move the temperature blend door.) All of those are inside the cabin, under the dash, if that's where the noise sounds like it's coming from. The other possibility in a Gen 2 is the servo actuator for the coolant control valve, which is outside under the hood, closer to the left side of the car (the side with the big thermos under the fender), if that's where you think you hear it coming from. If I had to bet, from your recording, I'd be putting my money on the control valve. The HVAC servos are small motors and tend to sound more like you've got a mouse scrabbling under the dash. A bigger sound, back-and-forth like a washing machine, coolant valve out under the hood. That's more like what I think I'm hearing. Whichever one of those (well, four) possibilities it is, they all start doing that for the same reason. There is a position sensor on each one. The computer knows where it wants each one to go, so it sends voltage to the motor until the position sensor says it got there. As they age, the position sensors make poor electrical contact, and they send jumpy electrical signals back. The computer believes the position signal and thinks the servo has moved off position, so it sends volts to the motor to move it back. Of course that makes the position report jump the other way, and the computer thinks "whoops, too far!" and sends volts the other way to try to move it back again ... over and over and over. It is super simple to clean up those position sensor contacts, at least the ones in the HVAC servos, their cases just have a few clips and pop right open. A cotton swab and some contact cleaner (the stuff I have around is from GC and it's called Jiffy Bath, and leaves a light lubricant behind) and you're ready for the next fifteen years. More about that in this other thread. Now, I don't think I've heard of anyone opening up the coolant valve sensor to clean it up, as opposed to just replacing the thing. Maybe it's just a matter of nobody trying yet, but it also might be more sealed because it's outside, and harder to take apart. If you can manage to get into it (non-destructively!), the same cleaning will give it a new lease on life the same way. You could just let the things keep doing that for years following Ed's example, but what eventually happens then is from going back-n-forth-n-back-n-forth-n-back-n-forth all that time, the contacts will have worn clear through the tracks on the sensor: That was from Clas Cronholm's post just a couple months ago. Let it go that long, and then if you decide you're tired of it, the free fix for five minutes of your time with a Q-tip is no longer an option, and your remaining option is a replacement for Toyota's magnificent price.
This just in: mr_guy_mann has opened one up, and taken pics. Actual, off-the-shelf potentiometer in that one, not open contacts and traces on a gear as in the HVAC servos. Not the kind of direct access to get squeaky clean with a cotton swab. Still, might work to just shoot some Jiffy Bath-like stuff at the pot and twist the shaft back and forth a bunch. That's what I do with my 1970s-ish stereo when the volume knobs get noisy, and it gets the job done. According to mr_guy_mann, "all the screws broke off". (Looks to me like one screw didn't, and the case did.) So if I were going to try to get into one, I think I would break out the Lisle 62140 right away, and not try anything else. Maybe that pot can come apart. Does that look like an e-clip on the shaft? I also notice that the screw holes in that pot mounting plate are elongated, so there is probably a correct voltage-to-valve-position calibration to shoot for when reassembling it.
