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

Waymo

This site may earn commission on affiliate links.
Also, while "spending time with family" has become a cliche, there might be some truth to it. Being a CEO is a stressful job so it is entirely possible that after 5 years, Krafcik was looking for a change in scenery. So the truth might be a little of both, the company wanted different leadership and Krafcik wanted a break so the timing worked out.
Not so sure about a planned exit - usually CEOs announce planned exits several quarters before the exit. But looks to be cordial than acrimonious - but even those could be because of lack of progress (see how Balmer departed Microsoft).
 
  • Like
Reactions: mikes_fsd
@Bladerskb I think you worked for Waymo at one point. Do you have any thoughts on Krafcik leaving and the new co-CEO's? Do you foresee any changes in how Waymo does things?
I actually haven’t worked for Waymo at any point.

but to answer your question. Krafcik is a good man. However I firmly believed krafcik was a bad hire. I always believed that you need a tech guy in charge of any software company. I think things will remain at its current traj and then suddenly accelerate from here.
 
Last edited:
... you need a tech guy in charge of any software company...
Paranoid software guys aren't going to cut the mustard. You need someone who can take risks. AKA Elon. Paranoid software guys aren't going to take blood risk. They won't risk blood on their hands. I have spoken to people at Waymo and this is the impression I'm left with every time. If there is a chance of an accident they are reluctant to release. And there will always be a chance.
 
I suppose @Bladerskb already posted a link in the FSD Beta Videos thread, but I think it would probably belong here as well.

YouTube playlist stats tell me I've published about 16.5 hours of Waymo content (with hours more on the way), and that's a lot of boring straight-line driving to watch for even the most diehard of AV enthusiasts. So over a few weeks, I rewatched through my entire back catalog, added timestamps of all notable events where they didn't exist before, then threw all of them into a single easily searchable website. (Plus a Waymo FAQ and spreadsheet where you can view all my rides + associated stats)

Introducing: JJRicks.com (woohoo!)
All those looking for footage of Waymo unprotected lefts, disengagements (those are a bit rare), remote assistance interventions, roadside assistance appearances, construction zones, pedestrian/parking lot madness, and the like: start here

Enjoy!
 
I actually haven’t worked for Waymo at any point.

My bad.

but to answer your question. Krafcik is a good man. However I firmly believed krafcik was a bad hire. I always believed that you need a tech guy in charge of any software company. I think things will remain at its current traj and then suddenly accelerate from here.

Thanks for the reply. It should be encouraging then that Dmitri Dolgov is now co-CEO. He was part of the original Google self-driving project. Waymo now has a "tech guy" in the CEO role.
 
My bad.



Thanks for the reply. It should be encouraging then that Dmitri Dolgov is now co-CEO. He was part of the original Google self-driving project. Waymo now has a "tech guy" in the CEO role.

If by tech-guy you mean somebody who is not able to state a coherent thought, and only provides empty buzzwords and platitutes, then by all means keep supporting Waymo. In most other people the BS meter should be going wild after watching his interview with Lex Friedman.

No Andrei Karpathy here, by any stretch of the imagination...

Waymo firmly on the path to total irrelevancy...
 
It should be encouraging then that Dmitri Dolgov is now co-CEO
Is there any precedent for co-CEOs that worked well ?

I Know Salesforce tried it recently - but one of the co-CEOs left the company within a year.

and only provides empty buzzwords and platitutes,

Oh well - reality of Corporate America is that by the time anyone comes up the ladder to occupy leadership positions - most "tech guys" have turned into 100% managers. They didn't come up because they were good at tech - but were good at making other people deliver stuff (and managed their managers well).
 
Last edited:
If by tech-guy you mean somebody who is not able to state a coherent thought, and only provides empty buzzwords and platitutes, then by all means keep supporting Waymo. In most other people the BS meter should be going wild after watching his interview with Lex Friedman.

No Andrei Karpathy here, by any stretch of the imagination...

Waymo firmly on the path to total irrelevancy...

You are cherry picking a 13 mn clip out of a 2.5 hour interview to try to make Dolgov look bad. Here is the full interview:


It's a good interview. Dolgov is highly knowledgeable. Dolgov has a PhD in AI, was a post-doc researcher on Stanford's DARPA urban challenge team, was a senior research scientist on self-driving cars with the Toyota Research Institute for two and half years, was the head of Google's Self-Driving Project for almost eight years where he developed what is now the best FSD in the world, then become the CTO of Waymo where he served for four and a half years and is now co-CEOof the company with the best autonomous driving on the planet. He's hardly a slouch. To dismiss Dolgov is just silly.

Karpathy got a PhD from Stanford in convolution neural networks, was a senior scientist at OpenAI for a year and a half and then moved to Tesla where he was the director of AI for almost two years, then became senior director of AI, in charge of the team working on neural networks and AP. He's led the AP team for almost four years now.

Karpathy is certainly an accomplished engineer in his own right. And being the Senior Director of AI at Tesla is quite an accomplishment, especially at his age. But IMO, Dolgov has way more experience and expertise than Karpathy.

Is there any precedent for co-CEOs that worked well ?

I Know Salesforce tried it recently - but one of the co-CEOs left the company within a year.

Maybe having two CEOs won't work. We will see.

Oh well - reality of Corporate America is that by the time anyone comes up the ladder to occupy leadership positions - most "tech guys" have turned into 100% managers. They didn't come up because they were good at tech - but were good at making other people deliver stuff (and managed their managers well).

That's not a bad thing IMO. To be an effective CEO, you need strong leadership and managerial skills. Being good at making other people deliver stuff is a big part of the job. As CTO, he undoubtedly proved that he has those skills at making people deliver stuff. Probably a big reason why he got the co-CEO promotion. But Dolgov is also a tech guy. He led the Google Self-Driving Project for 8 years! He led the team that basically created Waymo's FSD from scratch. He seems to have both the tech expertise and the managerial expertise.
 
Last edited:
Off topic but why does the thumbnail for that video show the false and misleading graphic of the SAE levels?
Do you prefer the graphic from the NHTSA?

nhtsa_sae_automation_levels.png


https://www.nhtsa.gov/technology-innovation/automated-vehicles-safety
 
Last edited:
  • Like
Reactions: mark95476
Oh, what is it you don't like then? I thought the presence of the steering wheel or indication of "driver input" in Level 5 were misleading?

No, I don't like this graphic that was in the Lex interview thumbnail:

0*Q6TKFqzRcUhRXFw_


It is wrong on many levels (no pun intended). For one, the SAE levels are not a gradual progression of paying less and less attention. L2 still requires full driver attention. Both L4 and L5 do not require any driver attention. So both L4 and L5 would be "passenger". So it is wrong to only show L5 as "passenger". Also, L4 is also fully the machine's responsibility. Machine responsibility is not just L5. Also, the transfer of responsibility is not that smooth. L2 still requires full driver responsibility. And L4 and L5 are both full machine responsibility. And it says nothing of ODD.

I like the NHTSA one better because it shows a greyed out "driver" and greyed out wheel for both L4 and L5. So it correctly shows that both L4 and L5 do not require a driver or any driving controls. And the text is pretty accurate. It describes each level pretty well. It correctly describes how L3 can self-drive but may require a driver to take over with notice. And it correctly spells out that both L4 and L5 can handle all driving tasks. It correctly differentiates that L4 ODD is limited and L5 ODD is not limited. And it correctly mentions that both L4 and L5, the driver is optional.
 
No, I don't like this graphic that was in the Lex interview thumbnail:

0*Q6TKFqzRcUhRXFw_


It is wrong on many levels (no pun intended). For one, the SAE levels are not a gradual progression of paying less and less attention. L2 still requires full driver attention. Both L4 and L5 do not require any driver attention. So both L4 and L5 would be "passenger". So it is wrong to only show L5 as "passenger". Also, L4 is also fully the machine's responsibility. Machine responsibility is not just L5. Also, the transfer of responsibility is not that smooth. L2 still requires full driver responsibility. And L4 and L5 are both full machine responsibility. And it says nothing of ODD.

I like the NHTSA one better because it shows a greyed out "driver" and greyed out wheel for both L4 and L5. So it correctly shows that both L4 and L5 do not require a driver or any driving controls. And the text is pretty accurate. It describes each level pretty well. It correctly describes how L3 can self-drive but may require a driver to take over with notice. And it correctly spells out that both L4 and L5 can handle all driving tasks. It correctly differentiates that L4 ODD is limited and L5 ODD is not limited. And it correctly mentions that both L4 and L5, the driver is optional.
Well let's not debate graphics. There are lots of variations if you search for "SAE Levels". I doubt the graphic artist gets the final say anyway.
 
Dolgov has a PhD in AI, was a post-doc researcher on Stanford's DARPA urban challenge team,
Yes, he is kind of the old guard. Question is - does Waymo need more of the same or new blood. Looks like they went with more of the same. Or, see my speculation below.

The co-CEOs are temporary. The current CEO departure may have been relatively abrupt - so they didn’t have time to search for a new CEO. They will get a new CEO and the co-CEOs will go back to their earlier position.

One clue would be - have they found new people for those senior positions ? Or are the co-CEOs still in fulfilling those roles too ? Anyway, I don’t thing we will have this same setup a year from now.
 
That's too bad.

I was looking forward to Krafcik/Waymo putting Elon/Tesla in his place after their little twitter exchange a few months back. I hope new co-CEOs are going to pick-up the ball and run it for Krafcik.


Krafcik is leaving Waymo to pursue other ventures. The current COO and CTO will be taking over as co-CEO.

Twitter announcement:


My thoughts: I like Tekedra and Dmitri. As COO, Tekedra knows the operational side and Dmitri as the CTO, definitely knows the technical side. So I think they are a good match to lead the company. I am really hoping that we see a faster deployment of the 5th generation and a bigger push to expand to more areas.
 
  • Like
Reactions: diplomat33
That's too bad.

I was looking forward to Krafcik/Waymo putting Elon/Tesla in his place after their little twitter exchange a few months back. I hope new co-CEOs are going to pick-up the ball and run it for Krafcik.

Honestly, I think I would prefer if the co-CEO's stayed away from Twitter feuds and just focused on making Waymo's autonomous driving even better and expanding Waymo to more areas. Results speak louder than tweets.