In addition to radio, phone and navigation, have the ability to control climate, sunroof, windows, charge port door, lights, web URLs, etc. using voice commands.
With version 10, Tesla greatly expanded the list of available voice commands and made them more natural (i.e. you can often say them in different ways). This Google spreadsheet has the most extensive list we’ve found.
I'd like more voice functionality also. I sorta would like Tesla to work with Android Auto. I know the self driving stuff is built around the existing mapping and navigating architecture. But to write and read text messages and emails and display pix received, maybe work with Google. I guess the car would soon be able to open the sun roof as well as hide your beer and evade parking ticketers 😉
If Google can do it, Musk can do it with his OpenAI company. Or license one of the many NLP engines out there that people license. I agree that the phone dialer by voice is the worst. For a tech company they have to get this right in the age of Google Assistant / Siri / Alexa that all nail this part - users are accustomed to this just working - not acting like my 2000 era cars where you might as well not use it.
If Google can do it, musk can do it with his OpenAI company. Or license one of the many put their that people license. I agree that the phone dialer by voice is the worst. For a tech company they have to get this right in the age of Google Assistant / Siri / Alexa that all nail this part - users are accustomed to this just working - not acting like my 2000 era cars where you might as well not use it.
Reducing distractions is the great safety-in-driving drive of the last decade. While the focus is on cell phones, the same argument applies to complicated cars that require looking at the dash to operate them.
In the past, I would have argued that the best modern safety feature is the heads-up display. I personally wouldn't drive without it. But we recently purchased a Model X for my wife and I realized that Tesla cars have a bigger problem: a lot of features hidden in a complicated menu system that requires reading and touching a computer monitor.
The solution is a greatly expanded set of voice commands. Virtually everything in controls and settings should be accessible by voice command. The technology is a simple extension of the existing voice command system, but it would represent a huge step forward in safety and convenience.
I can't imagine why Tesla doesn't make this their #1 priority.
Google's seems to be the best. Alexa is the hands down worst. Personally, I find any tech to be both useless and aggravating if it doesn't work at the three 9's level. I'm retired IT, and if our products failed at the frequency of the consumer marketplace tech, we would all have been bankrupt and out of jobs. So, I don't care if Tesla does anything at all until they can, as stated above, hit it out of the park.
Even more than expanding, need to just improve the NLP. When I try to do voice dial, it gets it wrong a lot. Google understands me perfectly and I never have issues. Especially around people that have less common last names. Or allow it to learn from the corrections you give it.
In today's world with Alexa, Ok Google, Siri, etc and the Tesla positioning as a technology company having voice commands for many other functions like sunroof open/close, door lock/unlock, etc seems obvious. I would also be happy with the ability to have a limited set of verbal commands that get trained with a setup function to customize for each person's voice. I would also be good with using my google or apple phone as the front end to take the voice command and then communicate it to Tesla via an API. This would also be a great demo feature for Tesla sales.
I'm fairly certain Tesla didn't write the voice recognition for their cars. It's most likely coming from a third party. Furthermore, it's ability to recognize song titles and artist names already indicates the voice recognition engine needs no further enhancement. They just need to program the car to act on particular commands.
Voice recognition is HARD. Try out Apple’s SIRI for example. More commands could me more difficulty in recognition. The engineers need to prioritize what is important to be under voice control vs. the "might be useful stuff". Personally, I wait until I am parked or at a light to attempt any reconfiguration other then the most common used functions e.g., close the sun roof.