Please don’t put in extra bright headlights on cars. Stock LED headlights being to bright for other drivers is already a massively common complaint — and then we have people installing even brighter ones? Please don’t.
It is also illegal to use non-DOT approved lighting in the US. Was behind a jackass today with a receiver mounted accessory red light that was excessively bright and made it look like the brakes were applied.
For colours to look natural you need your white light to contain lots of different wave lengths. It’s usually measured as Ra. Artificially looking LEDs are easily 10x cheaper than photography grade LEDs. Also, this guy is probably paying taxes and handling stuff the proper legal way. If you order from Alibaba, chances are you’ll not be paying taxes. Plus if they offer a 5 year warranty, they probably need to keep some money around for repairs.
In addition to the all the other stuff, including light spectrum differences, you can't just trust that a "37000 lumen" light (cheap from China ...) is such a thing. Some examples of "100,000 lumen" flashlights that ended providing more like 2000 to 3000 lumens: https://www.youtube.com/watch?v=6q_0wxzClkg
It's possible, they exist, many such LEDs are probably manufactured in China ... but the legit ones are probably more expensive, and you may need a more recognizable brand to do some QA, and keep pressure on the factory to not slip quality or inputs.
Consider the cheap screwdriver included with the lamp in this story: unexpectedly, many were more faulty than the cheapest $4 screwdriver you'd find in any hardware store. The more stories you read about manufacturing stuff in China, the more you'll see very strange things. It's not about nationality or anything, it's an extreme kind of optimization. If you didn't catch it already, maybe you didn't really need what you thought you asked for ... they're just checking/optimizing
Yep lumens are additive (though your eyes perceive them logarithmically).
I don't know much about car headlights, but chatgpt says high beams are typically 25-45 watts, and assuming a generous 200lm/w that gives you 5000-9000 lm.
Roughly speaking, it's expensive because it's 50 lbs & tons of electrical components (that are much higher quality than $24 headlights).
Just to add context those are just dumb lamps and I acknowledge that the product here has a lot more features including IoT support and the ability to change Hue.
Is it the ability to change Hue that makes this expensive?
> GroqCloud will wind down over 12-18 months. They'll either get laid off or jump ship to wherever they can land. They built the LPU architecture, contributed to the compiler stack, supported the infrastructure, and got nothing while Chamath made $2B.
We’ve entered a new era. Big companies don’t need your startup. They only need your smart guys. Just those few guys. You keep the rest of your engineers and figure out what to do with them.
And lately, the answer has been, “wind it all down”.
This sucks so bad for most of their employees. But it’s a signal to the labor market:
Be very honest about what you are when you’re considering working at an AI startup. Are you an AI expert? Or a TF/Pytorch monkey? There’s an enormous difference between those two things. If you’re not the key guy, require a good salary up front. Because I don’t see a future where the “acquiring” companies start needing you as well.
Or... Maybe we should start to think about how we let corporations get bigger and bigger? What happens if an entity (read: company) becomes so valuable, that it is basically indestructible? Does it have the power to change politics to their discretion? And as such, also influence the legislative?
Unfortunately, the time to be having these conversations was 40 years ago.
But you know the old saying: the next best time is right now.
It is, sadly, a near-impossibility that we could get decent antitrust under the current administration. But if we techies, as a sector, were able to pull ourselves together, genuinely recognize that the level of consolidation we have is very bad, and start collectively advocating for real change—for something more like what Lina Khan was doing under Biden, reversing the Reagan-era shift to the intellectually and morally bankrupt Chicago School interpretation of antitrust, and going back to actually forcing companies to prove that acquisitions will be good for everyone else, rather than forcing opponents to prove that they'll be bad in very specific ways...
Then we might have a chance to make real change, over the course of the next few years.
Here’s the thing, this isn’t an acquisition. This is Nvidia hiring a few guys. What’s the government supposed to do? Tell those guys they can only work for one startup and they can never leave it? You can’t tell people what jobs they can and can’t take right now. (Probably not ever in the US given our history with slavery.)
Now, how do you stop that?
Companies don’t need to buy startups, they just need the expert that startup hired. Once they have the experts, everything else takes care of itself.
Point is, antitrust works fine. It just fixes the wrong problem. That was really the problem with khan. She never really figured out that the technologies have passed the legal framework by. She should have tried to come up with solutions to the problem of monopolizing intellectual capital rather than the problem of monopolizing markets or intellectual property.
No; they're supposed to tell Nvidia they can't hire those guys.
This isn't about the people being hired; it's about the company that's trying to headhunt them.
And, once again: robust, sane antitrust would never have let one company get as big, as vital to the tech sector, and wielding as much power as Nvidia does in the first place. If there were 8 companies that do what Nvidia does, letting one of them hire these guys wouldn't be nearly as much of a big deal.
> Point is, antitrust works fine.
Not the way we've been doing it, it doesn't. Like I said, that's the whole point.
Unlike other SaaS "acquisitions" of late, this will be not as straightforward closure of a subscription business.
The $1.5B contract with the Saudi Arabia is substantial and investors will want that monetized too, there are also existing DCs GroqCloud have in the ME region and also other spots around the world that are quite valuable for their hardware and power agreements etc.
Nvidia has CIFUS and other regulatory concerns and also don't want to compete against their customers be a neo cloud provider, the Saudis likely still want their DC build outs to proceed.
All this to say, the remaining parts while no longer as glamorous is still worth a lot and cannot be easily sold to big tech co. GroqCloud is more be like Nokia Technologies/ Networks rather be killed.
---
As a result, staff not part of the Nvidia deal likely have solid jobs and also now the opportunity to climb the ladder quickly now that a lot of leadership positions have opened up.
They are also going to have to be compensated higher in cash or poached by an upcoming chip startup as they are no longer tied to equity options vesting scheduled of a very valuable company (Pre deal Groq or now Nvidia).
In any scenario they will come out better of the deal, not as much they could have in a full acquisition yes, but certainly better than most engineers not working for a hyped AI startup nonetheless.
Hi, I'm currently a Senior Engineering Manager leading 5 teams (4 managers & 40 people) and looking for my next challenge in the US (GC). I have leadership experience that includes both people and product side and I'm very strong in system design, databases & microservices. My ideal role could be a mix of responsibilities between 50% people/product management & 50% technical (architecture & hands-on coding) but I'm open for anything between 100% management to 100% individual contribution
Also, happy to discuss any kind of roles (full time, c2c, w2, remote/hybrid/onsite) and can start by March 1st 2025.
We're a social travel discovery app that matches your interests with your friends for perfect recommendations.
Although there is still a little bit of AI in one portion of the app because that's how we started, our recent pivot has been focused on social mapping and review features.
Sorry, I never got a notification of your response but, Bingo! It's the same idea! We actually do restaurants as well! I'm using OSM at the moment, so you can technically rate and review any feature that exists on OSM.
Thanks. What sparked my curiosity is while I'm writing code even a small mistake in the most simple programs can cost me tens or hundreds of milliseconds whereas behind a simple flushing to disk command there is so much happening at so many layers all in sub-millisecond or at most in a couple of seconds.
I recently changed my car's headlamps to Chinese LED which claims to be about 37kLm and I don't know how much it is probably less than that.
Two of those lamps costee me around $24 on Amazon US (pretty sure under $10 in China).
What makes this $800+ ?
reply