I’m back from a short trip to the Bay Area and an annual meeting for one of my LPs, where many of my discussions centered around the state of the venture market and OpenAI/ChatGPT. You can certainly feel the energy change when discussion goes from how long will this last to what else can generative AI do. These mood swings define the nature of the startup market in general; one day things can be absolutely amazing and the next not so. On the latter, Tom Loverro from IVP sums up the current market dynamic quite well. If you haven’t read the 🧵, please stop and do so now.
This is also in line with my predictions issue from 12/30 - What’s 🔥 #322.
On the hope springs eternal side, I can’t tell you how many energizing conversations I had with founders about ChatGPT and OpenAI this week. Once again, my belief has always been that it’s a transformational technology, and if it makes sense to build it into your product because end users will benefit greatly, then by all means you should. But remember it’s a technology, and you still need to solve a problem uniquely and significantly better than what’s out there because the tech is not the moat.
Case in point: Intercom, one of the first SaaS 🦄 minted in 2018 has smartly built an efficient business and has not raised a dime since then.
and here’s a deeper technical dive 🧵:
I’m sure this announcement will bring a renewed sense of excitement to the Intercom team while also from a business perspective help its end users do more with less. I’m also sure every single late stage VC on the planet is reaching out as this is a great way to have a “Chat GPT” investment 🤣
Case in point #2: Microsoft keeps cranking - even though the product is not as slick as Slack, it has significant market share and just leveled up productivity.
Case in point #3 - Notion’s AI is finally out to market. BTW, notice similarities in all 3 products? Text summarization, auto complete, and idea generation/expansion.
We all need a shiny new object while the markets still wobble from interest rate hikes, and we sure have it with generative AI. For older startups, it can be a Renaissance to breathe new life into your organization as a cutting edge company, amp up your existing customer base, reduce churn, and reach a whole new audience like Intercom. It’s also the greatest opportunity in the opinion of many a new startup founder, and this is where I believe the trouble lies. There will be fortunes made by a few like OpenAI as an infrastructure layer, but thousands more will crash and burn chasing the next new thing.
Hope springs eternal, but tread carefully and do not forget that cool tech does not equal amazing business.
As always, 🙏🏼 for reading and please share with your friends and colleagues.
🔥 up for initial funding announcement for Hypernative which is creating a new category, CryptoSecOps, to stop zero day attacks in crypto. It’s also another amazing 🇮🇱 team we’ve backed with previous experience building cloud infra and selling to IBM and Microsoft Security research. Pay attention as they keep finding hacks before they happen!@Boldstartvc we ❤️ new category creation, esp. in cybersecurity + we see same opportunity in crypto security as we saw in our day one investments in pioneers like DevFirst sec @snyksec + AI sec/MLSecOps @ProtectAICorp Here's more on why we invested:medium.comHypernative — CryptoSecOps, stopping zero day web3 cyber attacks, welcome to boldstartAs a day one partner for technical founders, we ❤️ to invest in founders building 🏗 the future of enterprise infrastructure including…
Huge - HashiCorp open sourcing it’s internal design system and internal document mgmt system which allowed it to truly be remote first - built on top of Google Docs with workflows built inSecond, is Hermes. We have a document heavy culture (PRDs, RFCs, memos, etc) and we've invested in internal tooling to make it easier to author, publish, discover, and subscribe to relevant documents.hashicorp.comIntroducing Hermes, An Open Source Document Management SystemLearn about Hermes, an open source document management system created by HashiCorp to help streamline writing and document processes.
Speaking of Hashicorp - looks like Cloudflare is building their own Vault - from Matthew Prince, founder/CEO
Netlify buys Gatsby - must read for OSS founders - 8 year journey and evolution from open source developer tool for users to company to enterprise use cases to cloud platform and now acquisition - one of the OGs in Jamstack
To see how we got here, let’s start by rewinding back in time. Eight years ago, I was a front-end developer, and I fell in love with React. At the time, React was a tiny emerging framework for apps, but I wanted to build a website with it. So I took a week and built my own framework on top of React, called it Gatsby, published the code to GitHub and deployed my site on a service called Netlify.
Over the next couple of years, pull requests started to trickle and then pour in. A community started to form. I turned from a solo maintainer into the tech lead of a rapidly growing movement.
Having come from the Drupal world, I understood the needs of content sites, and I realized we’d need not just a rich frontend toolkit, but purpose-built cloud infrastructure to make the developer experience truly seamless.
So I worked with other Gatsby community members, including my co-founders Sam Bhagwat and Dustin Schau, to turn the open-source project into a company. With Zack Urlocker, Gatsby’s CEO, we’ve focused on serving the needs of enterprise teams building on top of Gatsby.
Future of serverless…
Building a Staging Environment for Data Teams (Orit Mansur Riskified - h/t DevOpsWeekly) - great overview of the data eng process and while similar to software development, data tables don’t have a simple version control system
Our development stack includes the following tools and methodologies:
We use Snowflake as our main database
Our ETL processes are stored as SQL and Python scripts
We use Git as a version control tool and manage our scripts in a Git repository
We use CI/CD processes as part of our deployment
We schedule and run our processes using Airflow
Our BI frontend tools are Tableau and Looker
Need people to label and train data - OpenAI has hired an army of contractors to make basic coding obsolete (Semafor)
💯 on down rounds - clean ones that is versus trying to maintain artificially high price with debt or structure - does no one any good
Good stuff, but CryptoSecOps, sigh. :)
Excellent edition and spot on tweet by Tom. Thanks for sharing!