EAP and FSD pricing give hints to likely release readiness

[TIppy]

A lower margin is still positive cash flow... meaning costs were paid in full and the purchase price for the car is higher than the costs to build it (hardware and all).

At the time EAP was promised to reach parity with AP1 in December... we know now that it did not happen. As far as FSD there's no date. It's the difference between a short term bet and a long term.

Yeah, but who paid for it? If the margin was lowered, Tesla did, until they make it up on autopilot sales.

Hasn't Elon said that FSD features should start rolling out shortly.

 

[JeffK]

Yeah, but who paid for it? If the margin was lowered, Tesla did, until they make it up on autopilot sales.

Here you're starting to mix up cost and profit. The pure cost of the hardware itself from Nvidia, Bosch, etc was covered by the consumer when they purchased the base model. Profit on top of this is either lower or higher depending on if the consumer purchases the upgrade. In either case the hardware is 100% paid for, if that makes sense.

This is the opposite of what happens if they purchase something that needs extra hardware added, like a larger battery, more powerful motor, rear jump seats, etc.

Hasn't Elon said that FSD features should start rolling out shortly.

Elon has yes, and a rumor be a Tesla employee, but it'll be limited... like stopping for stop signs and stop lights for example.

 

[TIppy]

if that makes sense

No it doesn't. If they make less per car with the added hardware, they paid for it, accounting b.s. aside.

 

[JeffK]

No it doesn't. If they make less per car with the added hardware, they paid for it, accounting b.s. aside.

If I make an object and it costs me $20 to make it and I sell it for $50 or I sell it with a software option enabled for $55, in either case my $20 cost to make it is paid for whether I make $35 or $30 profit. The hardware is paid for.

Cost Of Goods Sold - COGS

 

[sandpiper]

@kdday - furthermore in the "FSD will happen" camp is this. FSD is a two part problem as I understand it 1 - object recognition. 2 - path planning.

• Object recognition ie "stay in lane and don't hit that shopping cart" has been on a continued improvement path very quickly since January. Given its continual improvement let's assume they'll have that to an acceptable level within a year.
• Path planning / decision making - this is the logic/reasoning problem which they could have been working on for several years now in secret using software agents and/or cars on private courses. This is the FSD nut and probably has at least some coded logic along with some neural network based "learned" outcomes not specifically coded. - "You see two vehicles in opposing lanes - do you slow down, speed up, move to the shoulder etc."

It's possible that that the path planning coding is largely completed and what we are actually waiting on for some FSD features to get turned on is for "Tesla Vision" to reach a sufficiently accurate state of object/lane recognition that Musk decides to give the green light for some early/cautious/low-speed FSD feature releases.

I agree with your comments, except that I doubt that decision making is nearly as far along as people think. I think the whole autonomy issue is swimming in hype at the moment.

Object recognition and path planning (at least simple aspects of it that you see on the street now), have all tipped into the realm of practicality because of the convergence of deep neural nets / high performance GPUs and the availability of large amounts of online image data. This is the easy part of the problem.

Highly reliable decision making, especially in unique situations, will require contextual awareness - something people acquire over a lifetime. That's not something easily trained into a DNN. And it's no easier to hand code than it was 5 years ago. We're not going to have contextually aware computers anytime soon, and so true autonomy will have to take a different path: infrastructure intended for autonomy. Here, I'm thinking about:

- highly precise GPS and accurately mapped regulated "autonomy approved corridors"
- controlled / limited pedestrian access
- vehicle/vehicle communications
- "smart" roads, where usual signage is supplemented with RF signalling to cars.

I think we're at least 30 years from any sort of true L5 autonomy, simply because it will take that long to build out the infrastructure.

 

[TIppy]

If I make an object and it costs me $20 to make it and I sell it for $50 or I sell it with a software option enabled for $55, in either case my $20 cost to make it is paid for whether I make $35 or $30 profit. The hardware is paid for.

If you used to make an object that cost you $20 to make and you sold it for $50, and now you've decided to gold plate it and it costs you $30 to make who paid for the gold plating? If you still sell if for $50, you paid the $10 for the gold plating, not the customer.

 

[JeffK]

If you used to make an object that cost you $20 to make and you sold it for $50, and now you've decided to gold plate it and it costs you $30 to make who paid for the gold plating? If you still sell if for $50, you paid the $10 for the gold plating, not the customer.

$50 - $30 = $20 profit. I simply made less profit, the costs were still paid by the customer. I might later change my price to $75 to increase profits based on the value gold plating might provide the customer, especially if I think customers will go for that price.

Check out the link on COGS and calculating gross margins. It'll help.

 

[JeffK]

Highly reliable decision making, especially in unique situations, will require contextual awareness - something people acquire over a lifetime. That's not something easily trained into a DNN. And it's no easier to hand code than it was 5 years ago. We're not going to have contextually aware computers anytime soon, and so true autonomy will have to take a different path: infrastructure intended for autonomy. Here, I'm thinking about:

- highly precise GPS and accurately mapped regulated "autonomy approved corridors"
- controlled / limited pedestrian access
- vehicle/vehicle communications
- "smart" roads, where usual signage is supplemented with RF signalling to cars.

I think we're at least 30 years from any sort of true L5 autonomy, simply because it will take that long to build out the infrastructure.

So here's the thing, Tesla vehicles are driving millions of miles per day and occasionally recording data to send back home. I'd argue that you can get multiple lifetimes worth of data in extremely little time. You can get different roads, different situations, different weather conditions etc.

As far as the decision making process during driving, it's not impossibly complex. One your way home today try to keep track of every decision you make and why. Can a computer make those same decisions based on the same data you had? If a computer can respond quicker, does it really need to slow down a mile in advance of a pileup or can it slow down a bit closer? You might be surprised at how little thinking you actually have to do.

 

[TIppy]

$50 - $30 = $20 profit. I simply made less profit, the costs were still paid by the customer. I might later change my price to $75 to increase profits based on the value gold plating might provide the customer, especially if I think customers will go for that price.

Check out the link on COGS and calculating gross margins. It'll help.

That's why I called it accounting b.s. $50 - $20 = $30 profit. The customer has a ,presumably, more valuable object for the same price, and I have $10 less in my pocket. So who paid for the increased value?

 

[JeffK]

That's why I called it accounting b.s. $50 - $20 = $30 profit. The customer has a ,presumably, more valuable object for the same price, and I have $10 less in my pocket.

The profit is irrelevant. We were talking about who pays for the hardware itself who pays for the COGS. The answer is the customer.

Also remember that Tesla did increase the base price of the vehicle at the end of Nov 2016 by $2000

 

[TIppy]

The profit is irrelevant. We were talking about who pays for the hardware itself who pays for the COGS. The answer is the customer.

No. What's relevant is how much money do I have in my pocket at the end of the day. If I had passed the extra expense of gold plating onto the customer and sold it for $60, then yes they would have paid for it. Since I didn't pass it through, I covered the cost for them.

 

[JeffK]

No. What's relevant is how much money do I have in my pocket at the end of the day. If I had passed the extra expense of gold plating onto the customer and sold it for $60, then yes they would have paid for it. Since I didn't pass it through, I covered the cost for them.

Again confusing COGS with profit. COGS is paid for either way... if you wish to keep the profit the same, then yes you can charge more just like Tesla did. This doesn't change the fact that the hardware was paid for.
If what's in your pocket is less than you would have made in the past then that's lower profit. Yet, there's still no out of pocket costs... You aren't left hanging with an unpaid bill.

 

[sandpiper]

So here's the thing, Tesla vehicles are driving millions of miles per day and occasionally recording data to send back home. I'd argue that you can get multiple lifetimes worth of data in extremely little time. You can get different roads, different situations, different weather conditions etc.

As far as the decision making process during driving, it's not impossibly complex. One your way home today try to keep track of every decision you make and why. Can a computer make those same decisions based on the same data you had? If a computer can respond quicker, does it really need to slow down a mile in advance of a pileup or can it slow down a bit closer? You might be surprised at how little thinking you actually have to do.

I'd agree, except for the exceptional circumstances. What do you do when you hear an ambulance siren? Or see a clearly distracted 6 year old playing on the sidewalk? Or see the lumber truck up ahead dropping bouncing stones onto the highway? Or get stuck behind a snowplow that's kicking up all sorts of crap, making it nearly impossible to see until you can get up beside them? That will be fine as long as the system can turn over control to a driver. But that's not true autonomy.

You need contextual understanding to know how to deal with these situations. We'll need much smarter AI systems than what can be stuffed into the dash of a Tesla, at least for the forseeable future.

 

[TIppy]

Again confusing COGS with profit. COGS is paid for either way... if you wish to keep the profit the same, then yes you can charge more just like Tesla did. This doesn't change the fact that the hardware was paid for.
If what's in your pocket is less than you would have made in the past then that's lower profit. Yet, there's still no out of pocket costs... You aren't left hanging with an unpaid bill.

I didn't say it wasn't paid for. I just question who paid for it.

If I sold them one of my original items for $50, then opened my wallet and gave them $10 to have it gold plated, did they pay for the plating? For accounting purposes, yes, I can include that in cogs. Still I'm out $10.

 

[JeffK]

What do you do when you hear an ambulance siren? Or see a clearly distracted 6 year old playing on the sidewalk? Or see the lumber truck up ahead dropping bouncing stones onto the highway? Or get stuck behind a snowplow that's kicking up all sorts of crap, making it nearly impossible to see until you can get up beside them?

• What do you do when you hear an ambulance siren?
  • Look around for the ambulance... if I have cameras looking around at all times to identify an ambulance I wouldn't need to hear it, in addition I can't hear it if my music is turned up.
• See a clearly distracted 6 year old playing on the sidewalk.
  • A computer can identify children and also if they are moving toward the road. A computer can also react much faster than I could, if needed.
• see the lumber truck up ahead dropping bouncing stones onto the highway
  • Not relevant for lvl 5 driving
• Get stuck behind a snowplow that's kicking up all sorts of crap, making it nearly impossible to see until you can get up beside them
  • radar sees through snow, but if the path is not visible the car would not and should not try to pass. It should simply back off and wait. Computers aren't impatient.

Basically, there are simple solutions to most of life's problems. Humans tend to make things more complicated.

 

[JeffK]

I didn't say it wasn't paid for. I just question who paid for it.

If I sold them one of my original items for $50, then opened my wallet and gave them $10 to have it gold plated, did they pay for the plating? For accounting purposes, yes, I can include that in cogs. Still I'm out $10.

You are out $10 in profit. No matter how you look at it, for accounting purposes 100% of the COGS was paid by the customer to maintain your profit you simply charge more money (like Tesla did).

Not only did Tesla charge $2000 more in the base model, but if you wanted EAP and FSD you'd have to pay an additional $8000 on top of the $2000 extra. There was a single month and a half where this wasn't the case.

 

[TIppy]

You are out $10 in profit. No matter how you look at it, for accounting purposes 100% of the COGS was paid by the customer to maintain your profit you simply charge more money (like Tesla did).

Not only did Tesla charge $2000 more in the base model, but if you wanted EAP and FSD you'd have to pay an additional $8000 on top of the $2000 extra. There was a single month and a half where this wasn't the case.

You assume the $2000 was for the ap hardware, not just normal price escalation. I paid that $2000 before the ap2 hardware came out.

Accounting schmounting. Tesla paid for it, until they recouped it with ap2 sales.

 

[JeffK]

You assume the $2000 was for the ap hardware, not just normal price escalation. I paid that $2000 before the ap2 hardware came out.

Accounting schmounting. Tesla paid for it, until they recouped it with ap2 sales.

You mean Tesla lost potential profit... it was paid for in the base price as the hardware is considered COGS

 

[Reciprocity]

Tin foil hat time-

After some thinking, and some late night crying because Elon said absolutely zip about EAP/FSD at the Model 3 reveal, I've come to accept that Tesla is very far away from an FSD release. Here's why:

Think for a moment about the features that EAP promises vs the features that FSD promises. Disregard what is "likely" or "over-hyped", but just read the feature descriptions of each package on Tesla's website. Ask yourself which set of features would bring more true value to the owner in terms of benefit for the cost.

Then, notice that EAP is $5k, and FSD is $3k. FSD promises significantly more benefits for the driver, yet it is 60% of the cost of EAP?

Granted, you have to buy EAP to get FSD, but the pricing doesn't seem to make sense when matched up against potential features.

I believe this is a clear signal that FSD is a pipe dream and that they're over charging for EAP. If both actually existed, they'd likely price EAP at $3k and FSD at $5k.

FSD is really $8-9k because you have to have EAP first. Also EAP is at a premium because it is THE feature many buy the car for. I agree that Tesla could charge more for FSD on its own, but $9k after the car is purchased is a lot for a $35k base car.

 

[TIppy]

You mean Tesla lost potential profit... it was paid for in the base price as the hardware is considered COGS

Call it what you like. Tesla covered the cost, until they made ap2 sales.