mattturck

I don't think that's all the general consensus

Benchmarking is here: https://surrealdb.com/blog/beginning-our-benchmarking-journe...

The OG market map of Data/AI returns for its TENTH year (2,011 logos in total vs 139 back in 2012)

+ 24 themes we're thinking about in 2024, including:

* Is the Modern Data Stack dead? * Consolidation in data infra * Checking in on Databricks vs Snowflake * The Rise of the Modern AI Stack * Where are we in the AI hype cycle? * Experiments vs reality: was 2023 a headfake? * LLM companies: maybe not so commoditized after all? * What's interesting: AI agents, edge AI * Is Traditional AI dead? * Is Generative AI heading towards a plateau? * Is NVIDIA overvalued? * Open source AI: too much of a good thing? * Is AI going to kill SaaS? * Is AI going to kill venture capital?

Thanks both for a thoughtful discussion.

Just one important nuance -- VCs like me who come in early and stay with portfolio companies for 5-10 years are both on the "buy" and "sell" side of the market.

So yes, a slower VC funding environment does impact valuations favorably for VCs, for net new investments (the "buy" side), so I'm talking my book to some extent.

BUT net new investments is only a part of the job (<50% for sure) for a VC like me. Most of my time is spent working with my existing portfolio (sitting on boards, etc). And a harsher environment is bad news for my existing portfolio (and/or anyone's portfolio), and me (and/or investors like me) as a result. Like a lot of VCs, I've looked like a "genius" for the last couple of years as many of my investments became unicorns, with huge paper markups. Now it's likely that the pace of markups is going to slow down significantly. Perhaps we're entering a world of flat valuations - or even downrounds? No markups is not a good look for VCs - makes it harder to raise the next funds, etc.

So yes, in the long term, it's good for VCs if we're able to invest in (the right) companies in the 2022-2023 cohort at lower valuations. But that will take 8-10 years to manifest into concrete results, by the time those companies become very big. In the meantime, VCs won't have a lot of fun navigating a flat round/downround environment (if it does indeed materialize) with their existing portfolio.

Confluent, Kafka, and the real-time data opportunity: a quick analysis of the S-1 ahead of Confluent's IPO.

The ML, AI, and Data stacks are all in early days compared to the broader software stack. We were curious about how the small but growing number of public companies in these stacks have performed compared to the broader market, so we created a new public index. Introducing the MAD Public Company Index...

Jack Hanlon, VP Data, Reddit, speaks at Data Driven NYC

Julien Le Dem, CTO of Datakin (makers of Marquez) discusses the OpenLineage initiative at Data Driven NYC.

Congrats on launching this - great for the whole data industry

State of the union of the data & AI ecosystem.

Some thoughts and notes on the Palantir IPO.

Overview of the upcoming Snowflake IPO

Inspired by the issues at ScaleFactor:

* Using humans in AI is a feature, not a bug

* Whether you can “fire” those humans along the way makes or breaks the business

* Not all business problems lend themselves well to AI automation

Yes on both. The second point is particularly interesting, and counter-intuitive. There's something about the necessary learning curve to figure out how to use a Bloomberg terminal that makes it sticky. People end up viewing knowing how to navigate the terminal as part of their professional skillset. Of course, that probably only works as long as there's no serious enough alternative in sight.

Yes, agreed. But the difficulty here is that a lot of those niches are very small, in terms of number of users. Assume for example you're building a product focused on munis, which I don't think is a very well covered niche for now. How many potential users are we talking about here? 20,000? A bit more maybe? Yet you need to build a whole technology platform to serve them. Only makes sense (from a venture perspective) if you use munis as a beachhead to expand into another area, but as it turns out, the next area requires you to pretty much start from scratch on a number of fronts, because it is going to be so specific. So building a large business here requires you to address a bunch of different niches and hire specific teams for each - lots of venture money required, IMHO.

Great thoughts here, and funny story on Cap IQ. Speaking of which, Cap IQ is exactly an example of positioning away from the Bloomberg terminal (with a core initial focus on investment banking as opposed to trading). Unclear whether there's a big juicy opportunity out there like this, like there was when Neal Goldman and his co-founders started Cap IQ in 1999. And great points on #4 and #5, although there's a whole subsegment (underbelly?) of financial services companies which are more cost conscious and where $20k per year for one user is a real investment. Bloomberg is not going after those as aggressively as far as I know.