Interesting overview. And on who will win, on one hand, as you said, Waymo's platform is better, and they just need to continue focusing on developing it cheaper. But on the other hand, Tesla has the volume in that you can already buy a car with FSD (although, we're talking about taxis, not consumer cars, but Tesla has assembly lines ready to produce 100k of taxis). Hmm, hard call...
... just looked it up, Waymo's 6th gen hardware is estimated to cost $20k per vehicle. That's getting down there. And if they ever do even higher volumes, will probably go down further:
https://x.com/fchollet/status/2023522267846815874
Which did things like drive off roads into parked cars and killed people. I feel like if Tesla hadn't been selling "full self driving" for years, the trust would be higher "no guys, we promise it's actually full self driving this time for real now" just doesn't have the same ring.
On the rollout:
It's notable that Tesla is building their own rideshare while Waymo is partnering with Uber. Tesla has the vertical integrated benefit of getting all profits but Waymo doesn't have to handle new user acquisition they just peel off existing requests and can roll out faster.
Tesla might "pivot" later to selling the 2 seater cybercab as a cheap base to competitors like waymo. It seems like Tesla has an identity crisis. They don't know how to be a consumer car platform (cybertruck halting production) they torched their PR with the green base that made them successful in the first place. Now sort of feels like a startup trying to "pivot" to find product market fit instead of a mature, well run company with solid strategic vision. There are a lot of jobs on the line so I hope they figure something out.
Yeah, agreed about Tesla's identity crisis. Tesla (and Elon) are really making a lot of mistakes in the past couple of years and at least regarding creating a great EV, they feel like they have lost their edge. They don't look as good as they did compared to what's currently out there and they don't seem to be innovating as much. Think Elon isn't interested enough in the type of cars normal people buy anymore, or changing the world with a low cost EV.
And yeah, FSD has been misrepresented for a long time now, but yeah, what I've been hearing it is finally getting really good. It's not completely hands free, but it's supposed to be very usable, and keep reading in reviews that this is the one feature that sets Tesla's EV's apart.
... And actually, on Uber, only in Austin (where said you are) and Atlanta does Waymo use Uber. Yeah, I used to live in the SF Bay Area, and had used Waymo a couple of times directly from their own app. It was a very natural experience. And recently stumbled upon an article that even in Austin and Atlanta, Waymo is going to discontinue using Uber in 2028.
Maybe a bit of both sides in my opinion. It shows the dysfunction of the government and our healthcare system, but it was also practical and helpful to the those left out of the healthcare system.
And to add to this, for also properly dealing with the 2008 financial crisis. Just as a reminder to those out there, the strategy of recent administrations for crises like this was to cut federal spending in order reduce deficits, and many of the politicians at the time were calling for cutting spending. Obama listened to his economic advisors who told him to stimulate the economy and get money into the hands of the people. There were some rough edges to their stimulus spending but it seems to have worked and our recovery was the envy of most of the world. In recessions, stimulus seems to have now become the norm.
The stimulus was too small, and the recovery too slow. The economy didn’t reach full employment until 2016, with the last full employment period being the late ’90s.
Agreed, but compared to most European and Asian countries, who chose austerity, our recovery was much better as most of themhad multi-year recessions or stagnation.
And agreed, we known now that we should have gone bigger, but this was the first time stimulus like this was tried, and they weren't sure what they were doing. Also, have heard that at the time, there was a lot of opposition by many politicians to stimulus, so there was only so much they could do.
yeah, at least for me, it was one of the biggest risks the Obama administration took, bucking convention, but resulted in changing how we deal with many finacial crises.
Ah, maybe "recent" isn't the right word, maybe should have just said "administrations of the past 45 years." Because, was mainly referring to Reagan and his deregulation of the US economy with Reagonomics, and how this type of thinking became the norm afterwards, not just for politicians but for the American people. This is of course my opinion, but think the way a lot of politicians in 2008 reacted to stimulus when it was proposed backs this up at least somewhat.
...and I don't have anything against cutting spending. In fact, actually think Reagan did the right thing in using this tool as well as deregulation to end stagflation. But it seems like cutting & spending should be based on the situation and for 2008, spending seemed the right course of action.
Hmm, seems like the web partially agrees with you. Just did a search and yes, Reaganomics increased defense spending but tried to reduce all other government spending. The overall idea was small government (although in practice, this didn't happen).
Actually, Intel's fab's are becoming their crown jewel. The whole "5 nodes in 4 years," where they tried to catch up with TSMC pretty much worked, which is why the stock originally rocketed up by 5x a few months ago. They aren't far behind TSMC now.
But, they still don't have many external customers for the fabs. Intels fab revenu is something like a tenth of TSMC's. Although, it'll probably grow year by year if they continue to match the capabilities of TSMC's chip processes.
From what I read, Intel's recent chip production strategy was a miss, in that they focused on chips for laptops and not for the chips for AI data centers.
More specifically, they focused on making laptop chips like Panther Lake (Intel's latest and greatest) and they meant to make a lot of their money here, and the data center chips weren't focused on as much (I think these would be something like the Xeon 6, but not sure, but I read that they also could have boosted production of older data center chips). But, as it turns out, laptop sales are down because of the demand for DRAM is sky high from the AI data center build out, and not as many people are willing to buy a super expensive laptop. If they had forecasted more correctly, they would have focused more on making data center chips.
Also, their forward guidance for 2027 is lower than expected too, because it will take awhile to shift their production towards building more data center chips, which they are in the middle of doing.
But, you may want to check other sources because have only been somewhat loosely following what's happening with Intel.
Hmm, am no expert at Lean, but on one hand, since it's a programming language, guessing it necessarily means like all other programming languages, it has bugs.
But on the other hand, agree with you that the chances that one its bugs causes a solved proofs to be wrong is probably very, very low. Yeah, each version of Lean obviously must be run against hundreds if not thousands tests of existing proofs to ensure it's working correctly...
... Although, one thing that is a little worrying is that some mathematicians have said the Lean code generated in these OpenAI proofs is supposedly really complex and undecipherable. It might be pushing Lean to corner cases that are not well tested.
It can possibly have bugs. That said, the "core" of Lean is a somewhat "simple" core that does simple things to check the proof, so it's quite rare to find meaningful bugs
don't disbelieve you, but just looked at the reference language, it seems pretty full featured as a programming language, even with basic classes, polymorphism, scope, namespaces... But maybe most of the mathematicians who actual code with stick with the simpler language features? And also, didn't look to deeply, these could be very simple features that actually aren't very sophisticated.
I mean Lean has two things: a full programming language, and proof checker. The programming language is used to create either code, or proofs. But the "created proofs" are using a very simple subset (a bit like an "assembly language"), and only need somewhat simple rules to be checked
So as long as the "somewhat simple" core that checks to proof is correct, it doesn't matter if the "programming language" has bugs: either in the end the "proof assembly code" generated by the program is right and is validated by the core, or it's not, and you need to revise your proof, or to check for either a bug in the programming language (it didn't generate the proof you were expecting), or in very rare case detect a bug in the simple "core"
Ah, I see, that makes sense - and am guessing that most mathematicians (like most non-developers) who work with a programming language don't write "good" code and often ignore the organizational features like classes or namespaces. Yeah, could see how they really are just interested in the core feature-set of a programming language... although, have seen a bug in some high-usage low-level code before, which was greatly surprising. Still that was only one case in years of dev work.
I agree, great coding is art too, and not just the code itself, but very much the problem solving process. It’s sad the highest levels of cognitive ability humans can do are no longer needed (or not nearly as much)
Agreed, really enjoy solving tough problems. Like having some interesting enhancement that needs to be carefully and surgical added into a complex codebase. Or writing an algorithm that grabs a bunch of info, sorts it in multiple ways, and uses this info to generate a useful result for our API (am thinking of specific past situations that have done this).
What is not fun is the repetitive stuff, where you have to search a code base for the same change over and over again, but it's not something you can just do a search and replace. Am happy to have AI do this (although, even this, you still have to read through every change).
Agreed, it’s incredible and scary how natural and human it sounds. It’s gonna be pretty much impossible for almost anyone to tell if they’re talking to AI
People keep laughing this off, but population contraction will cause the labor force to contract, causing resources to contract, causing war in order to steal other country's resources. Better hope the promise of that AI robot workforce pays off.
... just looked it up, Waymo's 6th gen hardware is estimated to cost $20k per vehicle. That's getting down there. And if they ever do even higher volumes, will probably go down further: https://x.com/fchollet/status/2023522267846815874
Looked online, Waymo is really starting to ramp up it's services. They're now in 13 US cities (including Atlanta and Austin): https://support.google.com/waymo/answer/9059119#zippy=
And are planning on increasing the size of their fleet by a factor of 5: https://www.reuters.com/business/autos-transportation/waymo-...
They will have a huge advantage and seems to me like Waymo is going to win out, but can never tell.
reply