Hacker News new | ask | show | jobs
by xiphias2 1155 days ago
None of these were bad ideas.

Sundar can't execute because he doesn't know how to do programming and doesn't understand the details.

He would never pass an engineering interview at Google, or any other software company (especially not OpenAI).

8 comments

I doubt any CEO can pass software engineering interview questions. That's a pointless interview. No CEO is writing their own app. What he lacks is vision to convert random sparkly ideas into actual practical products and to control his company to focus on one variant of one product.

Google often has 2 of same ideas competing internally, which might sound like it encourages a healthy competition but most of their competitions have failed. Gmail can use dozens of new features which would be useful but they launched Inbox instead, an app with a confusion interface and nothing more. Duo has great video compression but that's it. They never used that in Meet or Hangouts.

Seems like they have gold but would rather use bronze.

Marissa Mayer could code and I don't think it had a particularly helpful impact on her job as CEO. CEOs do not need to be technical experts, and most highly technical people, at least in my experience, do not excel in management positions and tend to micromanage too much.

Whatever problems Google has, Sundar's technical background has nothing to do with it. Which either way with an engineering and science background and 20 years at the company is in itself more than decent.

Might be the exception to the rule, but bill gates, extremely technically proficient and an amazing leader. I've read multiple accounts of staff who worked with him saying it was astonishing the level of detail he knew about their projects.
Could probably add Eric Schmidt to that list.
That's a good point, I think he wrote lex?
It has nothing to do with coding. His failings are that Google's product management is horrendous. Sundar has completely failed to set a clear long term product strategy for the company and doesn't seem to be making any moves to fix it.
I don't think CEOs of software companies should be programming experts. It sounds cliche but I think what is missing with him is he is not visionary. Follower and not a leader.
Can’t be both cozy and risky. Need someone not afraid to lose a nine digit pay package. Innovator’s Dilemma and all that jazz.
The one thing that’s distinguished the place during my tenure is ever-decreasing norms and shared respect and ever-increasing distance, here, you see enabling jealous empire-building and subsequent directionless empire-defending, not engineering problems per se
To all of the downvoters and naysayers replying to the parent:

it's the difference between technical, engineering leadership and bean-counters or MBAs.

It used to be that we expected the leaders of tech companies to be technical (old time HP or even Microsoft. At some point that transitioned into non-technical people running software companies and I agree that does not sit well with me either.

That said, Google is not a software company or a tech company. It's an advertising company.

That's funny, Google Cloud Platform sure doesn't seem like advertising. I haven't done super extensive research on the subject or anything, but at first glance it seems closer to something like AWS, which I think is also not an advertising platform.
My company pays its monthly Google cloud bill using a Chase credit card, and we get maximal credit card points because Chase classifies Google as an advertising company. Go figure.
How were Allo and Duo not bad ideas, considering Google already had multiple duplicates of those?
Allo was instant messaging with users identified by phone numbers rather than accounts. That was something a lot of consumers clearly found compelling (Whatsapp, iMessage), Google had no product for, and would not have been something that could be retrofit as a feature in their existing chat app.

Duo was two person video calls optimized for mobile phones, and allowing calling via phone number. Basically Facetime and Whatsapp. Again, a hugely popular product category that they did not have a product in, and with a paradigm that was not compatible with their existing video call app.

That's not to say these were good ideas, or that the projects were well executed. They were me-too copies of other companies' products, and possibly without enough understanding of what made those products popular. But the "duplicates" criticism is just lazy, and makes it seem that you're just parroting memes and did not know anything about the apps other than their names.

Their stupid names and simultaneous introduction killed them.

This is classic Google culture where Director A and B decide to get into the same category of product and refuse to work together. Unfortunately there is no adult in the room to put them in the place.

Tbf every company suffers from this. Look at multi tasking on iPad as an example. Classic Jobs might have been the only one who could cut this out

but surely it would have been easier to add "call contact" to Hangout/Meetings than create a separate app for the specific feature? It's not like you have separate contact lists, and I mean, that is how it works now[0].

So, no I don't think this is a lazy criticism, the fact that a new app was created instead of adding a feature to something that already existed is a managerial mistake because it creates extra confusion for users weakening the existing products[1].

[0] https://www.youtube.com/watch?v=uVFda0I8uW0&t=1s

[1] Sometimes it makes sense to extract a smaller app from a big one, but this is not the case.

Allo was basically an early (and of course much worse) version of what ChatGPT is now, it was a really cool chat bot, but most of the stuff was dumbed down totally and taken out by legal.

What you saw was not the original project that the engineers were trying to launch.

Is OpenAI interview considered particularly hard?