Having more than one type of sensory feedback simultaneously as an alert could prove incredibly useful, and smarter cars would have the capability to allow more advanced types of feedback. (Potentially as these might even be slightly customisable, although I am not proposing this as a core feature or prerequisite merely a consideration).
For example, supposing someone hooting also sent a light tactile vibration or pulse through the driving or riding apparatus from the direction of the person hooting. A smart car could for example facilitate this similarly to how the vibration on a cell phone would work (and as we know these can be very uniquely customised to very specific patterns in order to distinguish different types of messages).
Even better, a smart mirror could highlight digitally which car behind you was hooting through a digital outline or glow on the specific vehicle in the mirror, so you could know instantly where the sound was coming from, which would be essential if they are actually trying to get your attention towards them directly. Perhaps this slight glow could be colour-coded so you could feel the vibration feedback near the back of your chair and seat and to the right if that is the direction of the hooting car, and your mirror would simultaneously outline that car in blue as it hoots, then that glow would slowly fade. If a second car directly behind you hooted you would receive the feedback directly behind you and that car would be highlighted in a slight green glow/outline on your mirror. Then you could easily distinguish that two people behind you are hooting. If multiple cars are hooting you would see each of them lighting up and if all the hooting is coming from behind you you might deduce that something is wrong with the back of your vehicle (or you are driving below the speed limit in the fast lane).
For another example, potentially a smart HUD windscreen could highlight potential hazards on the road including when a car is approaching an intersection too quickly and has not started braking for a traffic light. Given self-driving cars would already be making these calculations, it would make sense to let drivers in on this visibility in manual mode, or co-driving mode, or in general for extra accountability of the self-driving system and safety for the passengers. Though this last one would not be designed to replace the hooting function, as passengers gain more trust in smart car systems (and hopefully fewer people opt for driving and we all become safer with established self-driving car technology and less human error) there will be fewer occasions for hooters to be used in general.
The benefits of this would be:
Accessibility and customisability for people who have impairment in one or more of their senses.
Useful to the general public when driving through poor visibility (eg fog) or audibility (in noisy traffic for example) or tactility (when driving through construction or somewhere where heavy vehicles and other strong vibrations are present).
In case one form fails to sufficiently alert the others are also present. We already have this in car systems having a combination of warning lights as well as sound feedback and tactile feedback - not only hooters but also various sounds and tactile feedback an engine/gears/brakes makes which are healthy or unhealthy, the click of indicators as well as the flashing lights on the dashboard. We don't currently have simultaneous multi-sensory alerts for hooters, flashing lights and hazard indicators but we do have these separate warning systems, showing that we recognise we may need to signify emergencies in different ways, even if not simultaneously.
Given we have multisensory feedback/alerts with success on so many of our automotive systems it’d be worth looking into how we could implement this for warning alerts as well.