A lot of my focus us upon artificial intelligence and all that is taking place. We are in a time where a lot of things are happening. If we do not focus upon the digital platform business model, we are going to miss out. This will be at the core of most successful entities going forward.
There are a number of companies that I watch regarding this. One is Amazon which is probably the poster-child for this. Another is Tesla who is starting to really excel in this area.
Sadly, most do not understand what is taking place. How many times have you heard Tesla is just a car company? That is like calling Amazon an online retail outlet. When I hear this, it makes me wonder if the person really understands technology. That is a dangerous place to be in this era.
On today's earnings call, Elon touched upon the different aspects to what Tesla is creating.
Here is what I posted on Threads:
https://inleo.io/threads/view/taskmaster4450le/re-leothreads-2qfk1zgez
Some might dispute what is taking place. However, like the Amazon naysayers for so many years, the reality of the situation eventually set in. While Amazon is one of the largest retailers in the world, most of its profits comes from something that was, at least to start, excess. I am referring to computing and what we know as Amazon Web Services.
For Tesla, there is a myriad of components feeding into the situation. That said, there is one area we will focus upon, Full Self Driving.
Many are aware that Google has its own autonomous driving division. It is named Waymo and it is an actual service. There are select cities where a Waymo vehicle operates similar to another other ride service.
In doing a quick search, as of the end of 2023, Waymo has driven a total of 20 million miles autonomously.
This became laughable when Tesla posted this chart a while back:
https://inleo.io/threads/view/taskmaster4450le/re-leothreads-fjsdexrf
I do not recall the date of the release. My thread was 17 days ago and the chart might have been out a week or so. Let us call it the end of March/early April for the total.
To start, we have 1 billion compared to 20 million. This is not even a comparison.
If this were the end of the story, it would be bad enough. That said, it gets worse from a competitor standpoint.
This was posted in the earnings packet released by Tesla.
If we open the image, we can see the total is nOW around 1.25 billion. This is up ~ 250 million miles in the last 25 or so days. This is more than 12x the lifetime total of Waymo.
Some might contest this focus upon miles. After all, this does not guarantee autonomous capabilities.
That is true. However, there is something that we have to consider.
AI basically boils down to data wrapped in algorithms and fed through massive amounts of compute.
A couple months back, Elon Musk along with a few others from Tesla posted on X that the company is no longer compute constrained. Another image from the earnings' packet tells the story there.
Notice what happened in the first quarter. The company went from the equivalent of 15K GPUs to over 35K. This is a massive increase.
We can now see why the company was willing to open up the trial to a larger group of people. It also made pricing adjustments on both the purchase and subscription for the software.
So basically we have a massive increase in the data along with the compute. The neural network is continually being trained on the new data collected, with use cases being isolated.
The flywheel, which is usually associated with revenues, can also apply to data.
While what is shown here is impressive, we are basically dealing with two countries. The United States and Canada are the primary feeder of this data. The company has not entered Europe or China with these trials. That is something that will change in the future.
Obviously, there could be regulatory issues to deal with. Nevertheless, there will be an expanded offering at some point, moving up from the 1.5 vehicle potential towards the 6 million on the roads globally.
Data is the infrastructure for all AI. This is what everything is built upon. Of course, that can be just raw data that has now labeling. In most instances, that is of little help. However, when a company is pulling vast amounts of data, labeling it as it comes in, and then training its system(s), things can happen rather quickly.
For Tesla, the trial of FSD, from a mileage perspective, was a huge success. There is a robotaxi unveil on August 8th of this year. That is a little over 3 months away.
The point here is Tesla is seeing the capabilities 3-6 months ahead of the general public. That means what they are training today is far superior to what is on the market now.
It is time to strap ourselves in. Things are going to get very interesting over the next 12 months.
