Apologies for the short preamble this morning as I’m writing from a 2 day AGM hosted by one of our LPs at boldstart. All I can say that it was quite a crowd, global in nature with both established VCs and emerging managers along with the LPs who invest in these funds. While AI was a topic certainly raised in many a conversation, the one that LPs brought up equally as much was around DPI (Distributed to Paid-in) which is a measure of how much cash a fund has returned versus what has been called down. In other words, have investors turned their paper marks into true cash returns.

The short answer is that 2023 was one of the worst years for DPI as there were very few IPOs in the tech industry along with a slowdown in M&A. While the LPs and VCs acknowledge that perhaps we could have all sold more on the way up, my view is this is also the time where steady hands 💎🤲 are needed. If you’ve done the hard work managing your burn and rightsizing your business, then rushing for the exit door when everyone else is will not generate the returns you are seeking.

Other interesting notes, thoughts:

• The quality of founders starting new companies is as high as ever and investment activity at the earliest stages has picked up significantly

• Series As and Bs still hard to get done, growth is still non-existent

• Tale of two cities for startups - many investors have seem to forgotten the ZIRP era as valuations for AI cos still defy gravity while solid businesses are getting less than stellar valuation multiples

• Smaller is better - funds smaller than $300M outperform larger firms although the larger funds are needed as massive asset allocators can’t write $50-100M checks efficiently into more focused, early stage managers

• There are over 5,000 funds <$250M 🤯 - we will see a mass extinction event for many small funds as well as LPs have limited room to add new managers especially if there is no DPI

• Multistage firms impacting the seed (inception market) in a big way and this is not changing any time soon which means that managers like boldstart need to have fund sizes to maintain ball control in any type of Inception round from discovery (<$2m) to Jumbo Megatron ($8-10 or >) and everyone acknowledges that first is still best place to be on cap table if one can do it efficiently

• Find LPs who invest for the long run as this could also be one of the best classes of startups to be built with AI tailwinds

That’s it - enjoy the rest of your weekend and if you want more free time, check out Superhuman’s latest release 😲. The simplest use cases are the most powerful in AI and this from @rahulvohra @Superhuman is a next level productivity enhancer 🤯

As always, 🙏🏼 for reading and please share with your friends and colleagues.

Scaling Startups

1. 💯

2. State of smaller VC funds vs. multistage $ Billion plus from Samir Kaji (🧵 here)

   Have spent a lot of time with institutional LPs (Endowments, Pensions, etc.) the last few weeks.
   
   It's clear that many absolutely would love to build a great emerging VC portfolio (and/or seed fund) portfolio.
   
   But only a small % will. Main reasons:
   
   1) Business model: Institutionals firms often need to deploy a large amount every year (especially large Pensions/Endowments), but cannot have too many line-items due to tracking and managing, it forces them into concentrating into fewer, but bigger managers. Many of these LPs have minimum tickets they need to write that are $10MM-$50MM (but not be too large a % of a fund). Rules out most smaller seed funds…

3. Deconstructing PLG from Mehdi Boudoukhane (Cycle - a port co) -

   The tech industry is obsessed with acronyms: AI, Web3, PLG, SaaS, you name it. But they serve venture capitalists more than they serve founders – the only thing they aid founders with is fundraising.
   

   The main goal of those acronyms is to put startups in glorified categories: "We're a [insert acronym] company" often means "we're a modern company, unlike all the other companies which are from the Stone Age."
   

   Not only is it bullshit, but I also believe that it misguides founders. Let's deep dive into PLG to illustrate the point.

   ❤️ the rant from Mehdi and reminds me of my prior post
   

   What's 🔥 in Enterprise IT/VC #361

   Ed Sim · September 30, 2023

   OK, can we now say it, the rage around product led growth (PLG) companies and the enormous valuation multiples relative to top-down enterprise sales cos was a ZIRP phenomenon. We now have post-ZIRP hangover and lots of cleanup for many a PLG 🦄 restructuring to go enterprise. Sure it’s an absolutely amazing mousetrap to acquire users top of funnel and c…

   Read full story

Enterprise Tech

1. AI will eventually “be used in almost every job” says JPMorgan Chase CEO Jamie Dimon from Miami 🌴 (CNBC)

   The burgeoning artificial intelligence tools from companies such as OpenAI still have their share of skeptics, but don’t count JPMorgan Chase CEO Jamie Dimon among them.
   

   The Wall Street titan told CNBC’s Leslie Picker on Monday that AI is not just a passing fad and is bigger than just the large language models such as Chat GPT. He compared the current moment favorably to the tech bubble around the start of the 21st century, when investor excitement seemingly got ahead of the actual changes.
   

   “This is not hype. This is real. When we had the internet bubble the first time around … that was hype. This is not hype. It’s real,” Dimon said. “People are deploying it at different speeds, but it will handle a tremendous amount of stuff.” JPMorgan has done work on the ability to use the new technologies internally, with Dimon saying that AI will eventually “be used in almost every job.”

2. Hype of magic? Read 🧵 - did Klarna’s AI really replace 700 customer support agents?

3. “The Surprising Tactics of Deepfakes Targeting Contact Centers Today” - Liveness detection, deepfakes, and four ways fraudsters use synthetic voices

   Brett Kinsela

   Synthedia and Voicebot have tracked the rise of deepfake technology for the past five years, and increasingly, the discussions revolve around the fraudulent use of the technology. Liveness detection techniques appear to be the most compelling solution to detect deepfakes and can be applied in real time or asynchronously to recorded media.
   

   Pindrop reports 99% accuracy against known voice cloning models and 90% for new models, known as zero-day attacks. Notably, both of these figures are for configurations that exhibit less than 1% false positive rates.
   

   False positives are an essential metric that is often ignored when considering deepfake detection. Any system could achieve 100% deepfake protection by simply identifying all inbound calls as deepfakes. If all calls are rejected, no deepfakes will proceed; thus, 100% identification is guaranteed. Granted, the false positive rate would be very high…

4. AI worms - be careful what you have AI agents doing for you? (Wired) - Security researchers created an AI worm in a test environment that can automatically spread between generative AI agents—potentially stealing data and sending spam emails along the way.

5. 👇🏼 from John Carmack, former CTO Oculus and creator of Doom - I’d say we are still a ways off from this happening, but once again, it does beg the question in the near-term, what happens when the >50% of code is written by AI? I wrote about the impact of this and some of these second order effects to consider in What’s 🔥 #379. Great discussion if you click on the tweet below…

6. One byproduct is more bugs and more bug fixing - this is pretty cool

   Also check out Testdriver from Dashcam (portfolio co), your AI QA agent, like an AI teammate that will do QA on your pull requests and repository changes

   

7. 👇🏼 Nikesh, CEO of Palo Alto Networks, caused quite the debate last week with his platform first approach and strategy to gain market share by giving products away for free for a period of time (also read What’s 🔥 #382). LinkedIn was mostly not kind when he went direct to the community. Here’s Nikesh coming back for more on LinkedIn doubling down on the need for a security platform. Once again the comments are enlightening and I’d click to read more.

8. 👀 The company said that it has moved almost 90 percent of its workloads to the Google service, after originally announcing a ten-year deal in early 2020…Still has some workloads on prem.

9. The Quest for a DNA Data Drive? (IEEE Spectrum)

   Moreover, DNA exceeds by many times the storage density of magnetic tape or solid-state media. It has been calculated that all the information on the Internet—which one estimate puts at about 120 zettabytes—could be stored in a volume of DNA about the size of a sugar cube, or approximately a cubic centimeter. Achieving that density is theoretically possible, but we could get by with a much lower storage density. An effective storage density of “one Internet per 1,000 cubic meters” would still result in something considerably smaller than a single data center housing tape today.

Markets

1. Another less than stellar outcome for a DevOps startup - Codefresh which raised $46M is now being sold for $28M (Fortune Term Sheet)

   Octopus Deploy, backed by Insight Partners, acquired Codefresh, a Mountain View, Calif.-based DevOps management platform, for $28 million.

2. Legendary CEO Frank Slootman retiring from Snowflake - end of an era (earnings transcript)

   Now, on the topic of CEO transition. I was brought to Snowflake five years ago to help the company break out and scale. I wanted to grow the business fast, but not at all costs. It had to be efficient and establish a foundation for long-term growth.

   
   I believe the company succeeded in that mission. The board has run a succession process that wasn't based on arbitrary timeline, but instead, looked for an opportunity to advance the company's mission, well into the future. The arrival of Sridhar Ramaswamy through the acquisition of Neeva last year represented that opportunity. Since joining us, Sridhar has been leading Snowflake's AI strategies, bringing new products and features to market at an incredible pace.

   
   He led the launch of Snowflake's Cortex, Snowflake's new fully managed service that makes AI simple and secure. Prior to Neeva, Sridhar led all of Google's advertising products. During his 15-year tenure at Google, he helped grow AdWords and Google's advertising business from $1.5 billion to over $100 billion. With the onslaught of generative AI, Snowflake needs a hard-driving technologist to navigate the challenges the new world represents.

   
   Sridhar's vision for the future and his proven ability to execute at scale made it clear to us as a board, that he is the right executive at the right time to lead Snowflake. This marks my retirement from an operating role. I will remain on duty as chairman of the board, and look forward to working with Sridhar and the team going forward. With that, I will pass it over to Sridhar.