The 4 Questions Startups Ought to Ask Themselves about Constructing with Generative AI by @ttunguz
There are 4 questions a startup ought to ask themselves about constructing a startup that makes use of generative AI.
I introduced these questions & my views on their solutions at Saastr’s Workshop Wednesday.
I had a blast placing this deck collectively. I began with just a few sentences, uploaded them to gamma.app to stipulate the presentation, popped over to Midjourney to generate pictures alongside the story line, & printed it in IA Presenter.
The video is here. The final slide incorporates the prompts for the pictures within the presentation.
The narrative is :
AI is a large platform change that Goldman Sachs initiatives will enhance GDP 300x greater than the PC.
GS estimates a 1.5-2.9% enhance to US GDP, doubling GDP progress, web of seven% job loss. The PC elevated GDP by 0.006%, in line with NBER
That alone ought to flip heads.
There are 4 questions startups ought to ask themselves about constructing with generative AI.
- Layer : utility, platform, or infrastructure? Within the cloud, AWS, Azure, & GCP have created about as a lot market cap as all the highest 100 B2B & B2C publics constructed on cloud (Netflix, ServiceNow, AirBnb, and so forth). However there are 100 functions in comparison with 3 infrastructure distributors.
- Market : how you can compete with incumbents? Startups have damaging time to launch in lots of markets with Adobe, Microsoft, & Salesforce launching Gen AI enabled software program in weeks.
- Moats : how you can develop aggressive benefit? Competing on algorithms is feasible however exhausting. Entry to proprietary information supplies a moat. Utilization & distribution, like in classical SaaS, are probably essentially the most sustainable & repeatable. Enterprise readiness can be a necessary : guaranteeing patrons are secure from authorized & compliance dangers.
- AI Depth : what stage of technical sophistication to convey into the corporate? Startups can combine with a plug-in, construct a prompt-tuning engine (somewhat mannequin on prime of a much bigger mannequin), or develop & prepare their very own fashions. Every subsequent alternative is costlier, however supplies a deeper moat. It’s probably startups begin at plug-ins & then transfer down with scale that affords extra utilization & extra capital to speculate.
When you’re constructing within the house,I’d love to hear from you.
