Product

aichannel

creating and destroying; flying cars, roc.camera, july.rocks

it seems like individual teams might care about their own product, but as you imply there is no urge at the leadership level to rationalize or make a coherent offering. their failure to translate gtalk into a viable whatsapp killer illustrates that its not just pedantic monarchic design brainworms, but actually affects their market position. their alphabet structure almost makes it work. like tencent's range of investments, which allow it to be everywhere without needing everything to be apple style centralized control. but they haven't permeated the alphabet architecture through google itself effectively. instead its a jumble of things that duplicate each other and work differently. might be more successful if independent teams were grouped under larger brands more effectively, the way youtube has performed very successfully despite having internal features that are duplicative of things elsewhere at google.

It's incredible how a product team's sense of what they're building affects their outcomes. 

Google Talk (Chat) is a classic example to me where they had the right brand, users, and scale to transition into a wechat/whatsapp style OTT messaging app, but didn't. They apparently saw themselves as building something else. Squandered their position. 

Similar dynamics playing out with Apple Intelligence — where Jony's new io product's positioning will make it able to do what Apple was well positioned to do with Airpods + Siri, but has failed to because they did not see themselves as building that product. 

Ambient audio AI has been an obvious next step for Airpods since their original launch in 2016, but according to reports in Bloomberg and interviews, it wasn't until they played with ChatGPT in 2022 that Craig Federighi and Jony Ive independently realized how important AI would be to UX. They underinvested in H100s and it will take them years to train competitive models according to Bloomberg sources.

director of product @seedclub. we built @crowdfund. i interview crypto builders @ web3galaxybrain.com.

very interesting take - case of too much ownership at lower levels without deep understanding of the business context for which they may be building the product..