Having worked in Failure Analysis you quickly learn that most manufacturers dont want to deal with some "known" issues. Those decisions were way above my pay grade
@billbell3737Ай бұрын
My friend and I worked at a semiconductor factory in the 70's. He was working on a problem where the fab yields would crater on occasion. They were looking at equipment, chemicals and procedures without any success. Over lunch one day we were discussing it and I suggested that it sounded like some person in the fab was performing some step that made the yields high and when he was on vacation or away that step wasn't being performed. He took that and did a manual spreadsheet and determined when a certain person was away it cratered. He was performing a second cleaning that he thought was useful and never documented it. He got a nice bonus. I can see where AI could figure this out a lot quicker.
@cratecruncher4974Ай бұрын
I used to mine OEM warranty claims data. I had to do it with spreadsheets and batch runs building a database. Filtering and correlating to discover patterns was actually fun. Even doing things crudely I found way more than we had resources to address.
@GG-si7fwАй бұрын
Your method is looking at data retrospective and the AI is suppose to be able predict the future, prospective. If your database can trend the future, that's awesome.
@ireneuszpyc6684Ай бұрын
@@GG-si7fw A.I. never predicted anything
@hughmackellar7941Ай бұрын
Excellent. As usual.
@technologyandsociety21CАй бұрын
As CTO of another generative AI company, I appreciate the value of this use case -- very cool!
@danneil8778Ай бұрын
Hi John and thanks.
@ashlyknapp1798Ай бұрын
now let's bring this to consumers live, fresh during the model year so they know what to avoid or develop remedies to avoid. I know how to pay for it.
@rwdplz1Ай бұрын
Will the AI tell the manufacturer all the H1B Visa Engineers they hired to try to save money is causing a record number of recalls and poor quality?
@teemutiainen860Ай бұрын
This should be used to analyse what is causing some Teslas to have multiple heat pump failures in cold climate while others have no issues. It seems this must be down to how users use their cars, but there are too many data points for a random user to understand. Times car started from extreme cold, times using fast heating, times car left outside, time driven in extreme cold, inside temperature used etc... Some users have had 4 failed heat pumps during warranty, while 95% have none.
@MossMiniАй бұрын
Legacy OEMs are parts assemblers and outsourced everything to do so.
@vancity2349Ай бұрын
Hi John suggestion for a video. In Canada most or all of the electrical grids are government owned. However in the US I understand many are privately owned and very profitable. This implies no incentive to expand or update build new etc. Thanks for
Ай бұрын
Great advertisement for Viaduct, who will Autoline be shilling for tomorrow.
@urbanstrencanАй бұрын
AI could definitely help with analysis for automotive market and cut down over production
@johngalt5411Ай бұрын
I always figured recalls & defects were a result of rushing product to market, quicker returns on product investment, etc. How will AI help with that?
@ashlyknapp1798Ай бұрын
from uniformity of communication through design, uniform material supply, consistent material mixtures, consistent assembly and consistent selling.
@slwiser1Ай бұрын
It sounds like they may be using Palantir software to discover this information. If not then Palantir would be a solution anyone could implement to do this same thing.
@ferfromlaАй бұрын
Not only the automotive sector but all manufacturing, could benefit. Boeing in particular could use this. In effect this system would serve as a whistle blower without the liability.
@JT_771Ай бұрын
The right data can be transformative. It's a big deal.
@AuralioCabal-nl8giАй бұрын
Maybe Toyota needs AI to help them with Tundra with Twin Turbo V6 recalls.
@mebran1193Ай бұрын
Ug. "That's a great question"
@dertythegrowerАй бұрын
As a user of Ai for a while, and part making.. I want to see it in action or it is another ai hype that (talks the talk but cannot walk the walk)
@gregkramer5588Ай бұрын
It is just a really good database analysis tool for this company. Any trend and correlations it can spot way quicker than we have in the past.
@talldave7799Ай бұрын
If he has another episode on this conference, he will have no choice but to quit the race! It's a big gamble specially after even Nancy Pelosi would not endorse his stay in the race!
@who2u333Ай бұрын
If this works as described, Tesla will be able to reduce physical vehicle issues quickly and easily. I am not sure that traditional auto makers have the ability to take advantage of this data and make on-the-fly changes to vehicle structure today.
@Derpy1969Ай бұрын
Generative AI should be able to also repair cars and point to the likely problem based only on fault codes.
@gregkramer5588Ай бұрын
That does not take generative AI.
@2012bigPermАй бұрын
And cuts jobs. All of our jobs eventually.
@dertythegrowerАй бұрын
another silicon valley 'Ai start up' like i said.. lets see it work or i am not buying it... and i operate Haas robots
@bobbbobb4663Ай бұрын
Pretty soon, babysmurf9000 is going to say Tesla is a AI Startup!
@federicotorresbordils7624Ай бұрын
If engines are designed by AI, should not be any recall, I guess......... BimmerN53
@gregkramer5588Ай бұрын
Completely different application. Very few recalls are from engines actually. But I am sure engine design is or will be addressed shortly. Many of the recalls are not due to the design but to mfg. issues.
@AuralioCabal-nl8giАй бұрын
@@gregkramer5588 I guess To wishes they had AI with their 3.4 TWIN TURBO Tundra debacle .
@highlanderapparelАй бұрын
To the Media team , that's watching This I'm gonna use chat g b t three point five Then I'm going to leap frog to chat GB.T five To bring all my family from facebook and I am gonna be setting up a lot of room Rooms on X, The Highlander😊