Welcome to Tesla Motors Club
Discuss Tesla's Model S, Model 3, Model X, Model Y, Cybertruck, Roadster and More.
Register

2020.4 Anything New?

This site may earn commission on affiliate links.
And I am pretty sure they could've re-allocate budget that pay for the UI dude's salary to hire neural net/FSD person instead. Assuming their wage cost is relatively similar.
Their wage would not even be remotely similar. Front End developer base is ~$70,000 and an AI/Machine Learning Engineer is between $140,000 and $240,000 IF you can find one. There aren't that many
 
Last edited:
Guessing you're HW2.5. As far as I'm aware this isn't an issue on HW3 (never been an issue for me on HW3). I'm amazed Tesla hasn't addressed the black screen bug in all this time. Bewildering, really.

I have never had the Black Screen issue on my car (July 19 Build, HW3), but it is not occurring on my car after this update. Started a few days after the update however.
 
Their wage would not even be remotely similar. Front End developer base is ~$70,000 and an AI/Machine Learning Engineer is between $140,000 and $240,000 IF you can find one. There aren't that many
And are you sure they only have just ONE(1) "Front End developer"? What if they have 10 and just reduce it to 8 and add extra 1 AI engineer? This is called resource reallocation.
 
And are you sure they only have just ONE(1) "Front End developer"? What if they have 10 and just reduce it to 8 and add extra 1 AI engineer? This is called resource reallocation.
Nope I'm not sure of that. But what is certain is you can't go to 0 front end (UI) developers if your software has a UI, changes elsewhere can introduce UI bugs that someone needs to fix. And at the scale of tesla we're not talking 10's of people on these teams. The scale is an order of magnitude over that. There's around 200 people on the Autopilot team alone. As well I think you missed what I implied. Machine Learning/AI/Neural Net Engineers are incredibly difficult to recruit. I work in the "Big Data" realm of Software engineering, meaning at this company and my previous company I worked hand in hand with the ML/AI/NN engineers. They are not plentiful this industry is a very small world and it's very difficult to recruit people. I don't think Teslas ML/AI/NN engineer headcount is a function of not having budget for any more .

In fact if you just look at their careers page they've had a number of long standing job postings in this area

There's also a Deep Learning infrastructure Engineer posting that's this year but that's more what I do that role would be less likely to actively push the state of Autopilot forward more quickly

you can't just pick these types of engineers up off the street
 
And are you sure they only have just ONE(1) "Front End developer"? What if they have 10 and just reduce it to 8 and add extra 1 AI engineer? This is called resource reallocation.


Also it doesn't take 2 full time UI guys to add games to the car.

It takes 1 of them. That's already on staff full time to do "real" work. And it takes him maybe an hour or two total to do it.
 
  • Funny
Reactions: groovyd
Also it doesn't take 2 full time UI guys to add games to the car.

It takes 1 of them. That's already on staff full time to do "real" work. And it takes him maybe an hour or two total to do it.

You are prime mgmt material right there, thinking porting a game to a custom hardware platform should only take an hour or two lol ;)
 
Nope I'm not sure of that. But what is certain is you can't go to 0 front end (UI) developers if your software has a UI, changes elsewhere can introduce UI bugs that someone needs to fix. And at the scale of tesla we're not talking 10's of people on these teams. The scale is an order of magnitude over that. There's around 200 people on the Autopilot team alone. As well I think you missed what I implied. Machine Learning/AI/Neural Net Engineers are incredibly difficult to recruit. I work in the "Big Data" realm of Software engineering, meaning at this company and my previous company I worked hand in hand with the ML/AI/NN engineers. They are not plentiful this industry is a very small world and it's very difficult to recruit people. I don't think Teslas ML/AI/NN engineer headcount is a function of not having budget for any more .

In fact if you just look at their careers page they've had a number of long standing job postings in this area

There's also a Deep Learning infrastructure Engineer posting that's this year but that's more what I do that role would be less likely to actively push the state of Autopilot forward more quickly

you can't just pick these types of engineers up off the street

oh you can alright... just you usually get what you pay for and sometimes you don't
 
You are prime mgmt material right there, thinking porting a game to a custom hardware platform should only take an hour or two lol ;)


Careful, your ignorance is showing :)

Tesla hasn't ported any games.

You think tesla wrote an atari emulator?

No- they took the off the shelf one that's been around for years and years and just loaded it on the car. This takes very near to 0 time. Creating the launch graphics and adding it to the menu would be the most time consuming bit.

Likewise cuphead, for example, was ported by the developers of cuphead
 
  • Disagree
Reactions: groovyd
Careful, your ignorance is showing :)

Tesla hasn't ported any games.

You think tesla wrote an atari emulator?

No- they took the off the shelf one that's been around for years and years and just loaded it on the car. This takes very near to 0 time. Creating the launch graphics and adding it to the menu would be the most time consuming bit.

Likewise cuphead, for example, was ported by the developers of cuphead

Ignorance is indeed showing. Who's it is that's for you to decide
 
  • Disagree
Reactions: MP3Mike
Ignorance is indeed showing. Who's it is that's for you to decide


Spoiler: it's you :)


News story about Cuphead on Tesla cars said:
Maja Moldenhauer from Cuphead developer Studio MDHR confirmed the work on the limited port to IGN's Ryan McCaffrey, saying Tesla only required that the game "play super, super clean" on the car's internal hardware.



Tesla talked to the game developer and asked the game developer to port the game to run on a Tesla.

So- as you've now been told 3 times- the only thing a tesla IT guy had to do was add it as a menu option on the UI. Which takes not NO time, but fairly close to it.

The atari stuff would've been even less work- since Linux atari emulators have been out there since mid 1990s.
 
  • Disagree
Reactions: Occar and groovyd
There’s a point here:
I’m not sure what it is, but I’m sure there is one.

Not saying they should rush responding to semaphores and stop signs as there is much more at stake there than TACC/Autosteer on controlled access roads but I’d rather they work on stuff we use while driving than stuff we *might* use while charging/parked..
 
... says every kid that's taken the TensorFlow tutorial these days... dime a dozen
Those aren't the people you want to be hiring. ML/AI/NN engineers that companies at the size of mine or Tesla ML/AI/NN engineers are mostly (but not always) PHDs. I'm often involved in the interview process for these positions in my company and I've been in this specialty for ~8 years now ( Data Infrastructure, I build and maintain the systems ML/AI/NN engineers use ). Actual qualified ML people are hard to find and don't move a lot. In any event if you want to discuss further we should do it in another thread or via PM this isn't the purpose of this thread.
 
Those aren't the people you want to be hiring. ML/AI/NN engineers that companies at the size of mine or Tesla ML/AI/NN engineers are mostly (but not always) PHDs. I'm often involved in the interview process for these positions in my company and I've been in this specialty for ~8 years now ( Data Infrastructure, I build and maintain the systems ML/AI/NN engineers use ). Actual qualified ML people are hard to find and don't move a lot. In any event if you want to discuss further we should do it in another thread or via PM this isn't the purpose of this thread.
tumblr_mc5w3fs5LV1rit1fvo1_500.gifv