Perhaps the best counter argument to dismantle the defensibility is to utilize pre-computed/trained networks (ala imagenet) that can be very accurate with a smaller dataset.
Показать эту ветку
Google’s relentless AI focus is now convincing every founder & VC that there’s basically no point in building anything that’s primiarily AI based. If you prove product market fit, they will copy, do it better, & kill you. This will force a new kind of schelp blindness towards AI
-
-
Спасибо. Твиттер использует эту информацию, чтобы сделать вашу ленту лучше. ОтменитьОтменить
-
-
-
Devil's advocate: What if they provide the tools to enable greater access to AI capabilities and enable founders to focus on product market fit and the best / most adaptive products?
-
Nothing would change. The theory is that they would only enable more ways to (for free) identify products that have found product market fit. Worst case they force you to sell with a threat of copying you & pointing their distribution towards it. Few exceptions will survive.
-
Sorta like Amazon and their Amazon Basics knockoff of products that sell well on their platform, only w/ AI tech
Конец переписки
Новая переписка -
-
-
Best bet is focusing on a narrow domain where cost of data acquisition is high and initial reward is low. Ideally you want an application with user generated data that's not openly available online, with a self reinforcing loop that helps you build a moat around accuracy.
-
Smart. +1
Конец переписки
Новая переписка -
-
-
Don’t agree. This is like saying if you have a product that runs on code, don’t bother developing because Google will do it better than you.
-
That’s the point my tweet: this perspective is taking over as the default & yet even if it isn’t true people will presume defeat before getting started.
-
Missed that sorry — in that case I completely agree with you!
Конец переписки
Новая переписка -
-
-
That word you use, "schelp"... it does not mean what you I think you think it means :) Schlep -- /ʃlɛp/
-
Maybe he means this: http://www.paulgraham.com/schlep.html
- Конец переписки
Новая переписка -
-
-
Any solution or suggestion?
-
It doesn’t need a solution. It’s a free market to attack Google or choose not to. Schelp blindness doesn’t favor the weak, fortunately.
Конец переписки
Новая переписка -
-
-
Haha but then you decouple their comments from what they’re doing... SPOTTED 5 minutes ago: Google employee (you can tell by hardware) walking around with a backpack to help Google win the AI race. Apparently Google Map cars didn’t do the trick; so add walking humanspic.twitter.com/NXZP4UskxG
-
What a sad looking mother f***er
-
When you show up to code and they hand you a backpack and tell you to start walking; Smh
- Конец переписки
Новая переписка -
-
-
AI is a moat like anything else. Just like web search content was a few years back. Folks will catch up and one of the big players will open up the gates to compete against them. there is still hope until regulations catch up. After regulations, monopolies solidify.
#GDPRСпасибо. Твиттер использует эту информацию, чтобы сделать вашу ленту лучше. ОтменитьОтменить
-
-
-
I don't think it's as simple as that. There are various factors. Sure, you won't get funding for doing an image classifier but there are lot more industries to be impacted which Google prolly wouldn't delve into.
Спасибо. Твиттер использует эту информацию, чтобы сделать вашу ленту лучше. ОтменитьОтменить
-
Загрузка может занять некоторое время.
Вероятно, серверы Твиттера перегружены или в их работе произошел кратковременный сбой. Повторите попытку или посетите страницу Статус Твиттера, чтобы узнать более подробную информацию.