Hacker News Reader

There was a recent post on here (https://news.ycombinator.com/item?id=41239520) where someone noted that people at Google tend not to use their own products. When it was posted on X a Google Director said that they use Google software because it's a software company. Which does show part of the dysfunction - you have a lot of hardware being built by people who aren't using it.

I have no idea if similar things happen for many of their software products, but that would explain a lot behind things like GHC and how it's UX has gotten worse over time.

There might be some truth to this, but under Ballmer's leadership, Microsoft's inflexibility re: the use of competing hardware came with its own problems. Since Nadella relaxed this policy, it's unclear to me whether you could say HW quality is worse. In any case, the significance of this policy may be overstated one way or the other.

> amazing quagmires built by geniuses

You’re describing Google 10 years ago.

My recent visits to its campuses have left me unimpressed. There are plenty of geniuses. But there are also many morons, and the mix is maybe a degree or two away from median corporate America.

>You’re describing Google 10 years ago.

The Google of 10, or 15 years ago did not build quagmires. Google's interview process is so rigorous that I do not doubt the caliber of talent that even currently works there. I think people don't have enough perspective when they criticize something like the Pixel Phone or Gemini. I don't think those are products that could be built by "morons".

There is a lack of risk or commitment to whenever Google launches anything new, and they refuse to be anything other than a fast follower. Stadia was an great product - afaik it worked with acceptable latency and the hardware was good. Morons could not have executed it that well. The quagmire came from the fact the messaging was confusing and there was no leadership that had any vision or conviction for the product. When I look at something like Stadia, as a microsm of new Google projects, the last lesson I take away is "Stadia was a shoddy product because people didn't spend enough time in the office." More engineers sleeping overnight at Google wouldn't have saved Stadia. Replace Stadia with any half baked release out of Google.

I worry the same for Waymo - an amazing product, damn near futuristic at this point, that will fall on the sword because no one at Google will be willing to risk their career on it.

>Stadia was an great product - afaik it worked with acceptable latency and the hardware was good. Morons could not have executed it that well.

Stadia was early in the game, cloud gaming will maybe take off in the next 10 years....it's still too early.

Google has brilliant programmers but I think product managers and business managers are the problem.

Maybe they just need to invent LeetCode interviews for PM and BA roles

> Google's interview process is so rigorous that I do not doubt the caliber of talent that even currently works there

I do! The process is rigorous. But in my opinion, rigorously wrong. It tests for problem-solving skills. But not something else. (I openly admit I don’t know.) It’s the freshman dorm commons at an Ivy League with a lot of money.

This nonsense clears out above the 8th tier of whatever ranking system they employ (principal?), for the most part. But it’s a lot of making increasingly-elaborate artisan firing mechanisms for guns that shoot a Petri dish of cancer, in the name of curing cancer. It’s technologically marvellous. But it’s so stupidly useless. (Bloom! Self-piloting boat that could mostly keep pace with the winds. Why? The balloon tore itself apart if inflating with even a bit of wind on launch.)

I think Google's average employee is still almost certainly above the average for the industry as a whole, but they're definitely now only "above average".

Google's big problem is that they've always been dysfunctional. The AdSense money printer let them grow the company 100x over the last 15 years without the company actually figuring out how to scale up or solve any of their structural problems.

It's also very strange to throw the average worker under the bus for what is clearly a leadership failure at practically all levels but obviously coming from the top. Google has a ton of innovation; the current AI race has deep Google roots but needs to capitalize the opportunity. Google needs leadership change, Sundar is no Satya and while he's done reasonably well, the market is too aggressive to let that retain leadership.

Google invented very little AFA actual products. They invented mind sets to make copying existing things with one questionably clever idea and turning it into the origin story of the world.

Exactly - GOOG product disfunction outside ads/search predates WFH by.. forever.

Nothing new here, just a new excuse by one of the guys that presided over it.

> you only need to look at Zuckerberg and Meta,

As much as I don't like what meta is doing in certain areas of their business, I think Zuckerberg has a pretty good corporate structure. From talking to colleagues and interview series I've been to there, it's a very "engineer-led" culture. Managers, at least a few years ago, were allegedly not even supposed to call it "their" team, they were supposed to say "supporting this team," etc. Very small nuances like that can actually cause a big shift in culture. It's also very flat, from what I understand. The most dysfunctional corporate environments I ever see typically have 5+ management layers above the lowest person on the tech rung. To me, that's ridiculous. It's just managers managing managers managing managers managing managers all the way down to the poor guys at the bottom trying to figure out what the heck is going on - it's impossibly anti-productive and prone to corruption.