Building with AI | Writing, experimenting, learning in public
Boston, MA
Joined December 2011
- Tweets1.5K
- Following129
- Followers325
- Likes15.1K
David retweeted
I also like to see founders self selecting into an area they’re passionate about vs me telling them where they could go and make a lot of money. The mission alignment will keep them going on a bad day!
interesting discourse happening around email deliverability and build vs buying saas. I didnt realize how hard deliverability was until this conversation started about mailcheap vs beehiiv.
from a builder perspective its a good lesson in nuance and touches on product market fit... as in deliverability is one of the problems these products solve and thats hard to vibe code
Proudly using postshiba.com/ by @jessethanley to send emails (not AWS SES).
The man has built an email ecosystem packaged in a super cute Japanese design. I'm reaaaaly envious of his creativity and skills 🥹
(not affiliated)
all this AI paradigm shifting is low key giving me an existential crisis
it’s not knowing what to work on because of the rapid advancement of AI
coupled with the feeling of (frankly) grief from coding being automated / feeling rug pulled
Let me summarize a few convos with founders from past 2 weeks:
Personally, I won’t invest in an agentic “SaaS” / B2B company anymore no matter how good your AI and agent is …
Unless it unlocks materially more budget than just traditional software budget.
This may sounds obvious but I think a lot of folks get confused by it. It’s not enough today to use AI to make your B2B product 10x better than the old guys. That’s great and it WILL win you customers.
The problem? That alone won’t win you incremental budget.
You can steal Salesforce’s budget, or Workday’s, or whatever, but all the incremental budget is going to AI that either has radical ROI (gets you customers) or radical cost savings (don’t need half support team) or radical insights (100x better analytics -> insights -> ROI). Agentic ROI.
Stealing budget just from other software is no longer enough.
-> You have to have Agentic ROI, and where that shows up is much larger deal sizes than usual in your category.
If deals are usually $20k in your category, but you can get $60k-$200k with AI, look, that’s the sign.
If deals are usually $20k and you can get $20k but you can be incumbents with AI, that’s great. But it’s not enough.
If your deal size isn’t at least 2x what your pre-AI competitors have, you don’t have Agentic Revenue, and you won’t scale to be bigger than them, and you may end up in … a 2x ARR trap at scale. No one wants that.
kind of want it to motorized so that it spins around to the person that’s talking, that would have given it some more wow factor
he said in the interview the review with Theo Vonn that he doesn’t do any substances even coffee. So I give him some credit for being polite and having a beer on the Cheeky Pint podcast.
general public asks chat medical questions and then argues whether it’s correct. that’s the lens they see AI through.
Not building SaaS, most don’t even know what that means.
the potential of VR will be realized when we can generate 3D on the fly using AI. Building 3D models of things is too slow when done manually so there’s not a lot of content on these devices but when it happens in real time it will be like magic.
Sam Altman: "The world has not caught up. 3 months of startup work is 17 minutes"
We cloned World Labs in 24h (they sold for $8.4B)
> text-to-3D, in minutes, environments where every object is a mesh and humans + agents can live and play together
Crazy times.
(more soon)
I'm down about 19 lbs thanks to AI
back in February I was about 270 lbs. I built an app to track my weight and what I eat. The app uses that data to estimate TDEE and set up a deficit.
everyone is saying that tou could build it with AI but not maintain it… doesn’t make sense
if AI can build it why wouldn’t it be able to maintain it?
