The AI battle is here..and customers will lose the war
who let the dogs out
3 things happened in the last few days
Slack is now banning third party apps (like Glean) to access Slack data
Post Scale.AI acquisition, Google (among Scale’s largest customers) will stop using Scale platform
Microsoft and OpenAI are having a bit of bad relationship. OpenAI is acquiring Windsurf - which competes with MS’ GitHub Copilot and MS isn’t happy.
[Quick announcement: This newsletter will be merged with BigIdeas newsletter soon - to deliver a whole lot of big ideas/a more streamlined curated experience - across text and audio]
Apart from the fact that data continues to be the new oil (and drone..and everything else), my bigger takeaway is that it will become very difficult for startups to compete when your ‘core’ relies on data stored somewhere else.
For e.g. what stops Gong to not share data with third-party apps - which in that case will pretty much make a lot of AI sales coaching companies quite irrelevant.
Companies who have data are hoarding data to train proprietary models and choke innovation - fair in a competitive world, but what about the customers?
Silos breed mediocrity
We all know this - the state of search engines before Google, the state of browsers before Chrome etc.
Silos tend to default to mediocrity.
When it comes to AI, businesses bet big on AI for efficiency, innovation, and growth. Instead, they get fractured ecosystems.
Slack’s data ban cripples third-party AI, leaving employees with incomplete search results.
The Microsoft-OpenAI feud fragments coding tools, disrupting developer workflows. ROI lags as integration costs soar. Like streaming subscribers paying for multiple apps to watch one show, customers face a bloated, underwhelming AI reality.
The war’s winners hoard data; customers foot the bill.
For e.g. Is Salesforce really the best when it comes to delivering value in Enterprise search space? No - but if they remain the only one, then customers will have to live with an average / mediocre experience.
Which means delayed and slow(er) enterprise adoption.
If you are a startup whose moat is built on someone else’s data, how are you looking at the future? Pivoting to consumer?
What’s your take?