What’s 🔥 in Enterprise IT/VC #447
Just another insane week in AI, what it means for founders + keep thinking bigger
Just when I thought Microsoft was killing it with its BUILD conference and then Google one-upped them with I/O, OpenAI came storming out of the gates, before Anthropic stole the show. Welcome to the another week of AI progress. Let me recap quickly and share what this means for you as founders/investors.
Google I/O
Deedy from Menlo Ventures summarizes all of the best posts covering coding agents, real time voice translation, generative video, and replacing designers 😲.
We, of course, had the viral post of how your startup will be eviscerated!
What’s amazing is that Google was left for dead, but when the founder comes back from the bench and goes into hard core founder mode, well you see the results.
Microsoft Build
Not to be outdone, Satya covers the big theme at BUILD - to be the platform where developers build the open agentic web!
Anthropic
OpenAI
If you’re feeling a bit exhausted keeping up with the latest releases and need some inspiration, watch this. It’s all about THINKING EVEN BIGGER and REIMAGINING what can be done with AI. Are we looking at the next Apple, Microsoft or Google? BTW, the X post has 14M views!
This is how ambitious Sam is ,and what’s at stake in the future. Not sure we even need another device at the moment, but this is damn good marketing selling the future!
Sam mentioned one of the reasons for the acquisition was all about TALENT DENSITY! This is a theme I’ve repeatedly written about even when it comes to much smaller acquisitions like the most recent one of my portfolio co Protect AI by Palo Alto Networks for $700M. The moat for startups is team especially in a world of dozens of competitors appearing overnight. Watch this clip from Nikesh this past week talking about VCs as his R&D lab and why Protect AI and the team!
Back to OpenAI - this is how you win big - fusing software, hardware, physicists, engineers and leaders…Essentially OpenAI paying $6.5B for the acquihire of a 55 person team per Axios.
The products are decades old - surely we can reinvent these legacy products.
This technology deserves something much better.
Founders, also think about your product, your workflow - are you bolting on AI or reimagining what is possible?
What this means for founders
As a founder, you need to keep asking yourself two things:
Does my product/service/workflow get better with each new model release or I am fearful of being vaporized overnight? Ok, it’s perhaps not that dramatic, especially as model providers continue moving up the stack, but you get my point.
Do I build for what models can do today and spend inordinate amounts of time jury-rigging my product to make it work now or build for the future?
How do I continue to become a talent magnet for the best folks possible?
Anthropic’s head of developer relations sums it up quite well.
Keep building, shipping fast, learning, and hire amazing talent - it’s still so early out there. What’s also clear is that the app layers gets more and more interesting every single day with agentic workflows still to be built to cover the last mile of the enterprise.
Since Microsoft is endorsing open standards there is are still some select opportunities to rebuild the infra stack for an agent first world. On the one hand every startup is dead whenever a large tech co announces 15+ new products, and on the other hand, these incumbents are just validating the market opportunity for all of us. If you agree on the latter, there is no greater time to build 🏗️.
Finally, what’s clear from all of this is that the biggest winner may very well be compute providers.
The demand curve is insanely steep, but we still have supply constraints!
Let’s keep building 🏗️!
As always, 🙏🏼 for reading and please share with your friends and colleagues!
Scaling Startups
#💣 "It's a Molotov cocktail of opportunity...just buyer beware."
if you’re looking to make sense of what’s happening from my vantage point in venture as an Inception partner to founders, at the idea stage, this was one of the most entertaining interviews I’ve done.
Peter Walker from Carta has some of the best data insights and we dove deeper here in this hour long conversation on whether we’re in a bubble for AI, how we think about inception investing, moats, exits, and more…
Listen here:
Watch here:
#❤️ this quote from NHLer Brad Marchand who by all accounts should have never made it - 👇🏼 clickthrough for the rest
Enterprise Tech
#Microsoft’s clear positioning which I 💯 agree - however the nature of the jobs may change to managing a team of agents but point remains
#best 30 minutes or 15 at 2x speed on our agentic future!
#betting on the future is super expensive - Perplexity in talks to raise at $14B valuation along with more from The Information
#$100M Inception rounds now becoming a regular occurrence 💰
#Palo Alto Networks revealed the acquisition price for portfolio co Protect AI - $700M 🤯 - more on why and the opportunity ahead! Also Elad Eraz has a great breakdown of the PANW quarter on this LinkedIn post
and the big 💰 slide $15B TAM!
#great reminder of how software gets discovered - agents! Also why Neon grew so fast and now being acquired by Databricks for $1B
more from my colleague Ron Miller’s post from FastForward
#how does one monitor no only long running work but also super fast work? Why I also believe we need a whole new cloud purpose built for agents - check out Agentuity, a portfolio c0
#Coding agents are the future with Devin, Codex and Jules but we have a problem - how do you not only review all of this code but also do it with guardails based on your own development policies, your own standards and conventions?
Welcome Baz agent (a portfolio co) - your custom reviewere based on YOUR codebase and PRs and finetuned by you over time
Give it a whirl!
#where are we with deployment of agents in the enterprise - survey data from PWC
#at last year’s JPMorgan Investor day, they talked about GenAI and this year revealed already >100 GenAI applications in production 🤯
#how JPM thinks about AI and ML when it comes to its $18B, yes $18B IT Budget
#we know Microsoft supports MCP but CTO Kevin Scott shares how seriously committed they are
#wow - 70% of Anthropic pull requests generated by AI
#we still have a supply constraints - need more inference compute!
Markets
#no man’s land
#regarding inference in my preamable is why cos like Fireworks AI exist and also have insane growth 📈
#as VC firms get larger and start bleeding into PE territory, always important to keep in mind how these folks are performing - supposed to be steadier returns with less risk as compared to VC
