theonlysandman

I will show up to chant "Gravy Plane, Garvy Plane" at the next Ford Event.

The list is longer than the song.

The deterministic spoon beat the probableistic spoon. 9 out of 10 soup eaters agree!

ROTFLOL

This is a fantastic visualization

I love this! 🤣

This is one of my favourite games of all time! So many different permutation to win.

Maniac Mansion (1987, Lucasfilm Games) is widely credited as the game that perfected and popularized the predefined verb/command interface in mainstream adventure gaming. While Infocom and Sierra games still used text prompts (with lots of trial and error), Maniac Mansion's predefined commands made gameplay easier and more intuitive, reducing the frustration level of pure text adventures. Instead of painstakingly trying to find the one correct term ore phrase to do complete a task, the predefined commands made gameplay more fluid and fun (at least for the casual gamers). Although a few earlier games existed (e.g., Enchanted Scepters in 1984 or Labyrinth in 1986), Maniac Mansion's mainstream appeal, commercial success, and quirky setting established the standard.

The world of software leaps forward right now _by the weekend_

@ameetshah Tom said the same thing! :-)

I want this, but for VC.

@JaredSleeper I doubt all their software will be built in house -- but for the best companies some of it probably will. Maybe the verticalized software will come from a HVAC company to other HVAC companies.

@theonlysandman Would be seriously impressed if the average HVAC services company builds all of its own software one day

Every day for the next long while, I'm going to tear down a new public software company and highlight the AI risks/opportunities around it- products launched to date, top startups, key quotes from earnings calls, etc. Day two, HubSpot: Peak share price: $841 (November 12th, 2021) Share price today: $232 (-72%) EV today: $ 10.9bn ARR today: $3.2bn (+21% Y/y) NRR: 103% EV/ARR: 3.4x GAAP Operating Margin: 1.4% EV/Run-rate GAAP EBIT: 243x Headcount: 12,234 (+21% Y/y) AI bear case: AI-native CRM/marketing automation startups that are natively built around agents as users disrupt HubSpot's core businesses as the value of HubSpot's UI dissipates. GTM moves to something more automated/mechanized vs. being a human endeavor. Seat count declines as AI automates sales/marketing jobs, which leads to secular pressure. AI bull case: AI-native CRM companies are still operating off of the same primitives (building GUIs, etc.). HubSpot has a broad platform (Martech, Support, CRM) which will drive customer stickiness. HubSpot's team is world class, including technical founder @dharmesh, and will keep the company at the cutting edge of agent adoption. SMB customers are much less likely to vibecode their tools than enterprise customers. AI traction: Non-disclosed, but business has re-accelerated for the last two quarters and is layering in AI usage credits. Adjacent startup summary: Clay- 1090 employees, +77% (not directly competitive) Unify- 191 employees, +195% Attio- 142 employees, +80% Reevo- 98 employees, +250% Clarify- 35 employees, +150% Rox- 76 employees, +192% Management Quotes: "As I shared last quarter, we expect to see a step-up in net revenue retention in Q4, resulting in a couple of point improvement in net revenue retention for the full year of 2025" "Customer agent now has over 6,200 customers, up 48% from last quarter with an average resolution rate in the 60s. Prospecting agents has been activated by 6,400 customers, up 94% from last quarter, and customers have used it to engage over 1 million prospects." "And specifically, with AI, we see the opportunity to monetize both through seats as well as credits. Now core seats, which we talked about at Analyst Day, it is becoming more valuable, and it is embedded with AI data platform value, and that opens up a large opportunity. credits gives us a new way to monetize usage for customers as they consume more AI and agent-driven actions." Additional Commentary: Prima facie, HubSpot was very well situated in a maturing SaaS market given its status as a consolidating platform spanning marketing, sales and CX. Given that, it's hard to argue that any one startup is effectively challenging the core business. Mid-market positioning makes it less likely that customers will internally build their own applications. The key question for HubSpot is whether white collar/marketing/sales work is automated writ large by agents in a way that elminates the need for software entirely, especially if HubSpot loses that market to an upstart.

@JaredSleeper Bear Case: Canary is trust: do lighthouse customers hand over outcome control? Or better question, to what extent with companies hand over outcome control?

Every day for the next long while, I'm going to tear down a new public software company and highlight the AI risks/opportunities around it- products launched to date, top startups, key quotes from earnings calls, etc. Day four: Braze Peak share price: $94.16 (Nov 19, 2021) Share price today: $ 17.44 (-81%) EV today: $1.8bn ARR today: $726m (+22% Y/y organic) NRR: 108% EV/ARR: 2.5x GAAP Operating Margin: -20% EV/Run-rate GAAP EBIT: N/A Headcount: 2,230 (+17% Y/y) What Braze does: Braze is an enterprise-scale system that lets companies trigger messages to users across channels like email, push notifications, SMS, and in-app based on real-time behavior. It acts as a system that collects event data, segments users, and runs rules-based workflows to decide what message gets sent and when. More recently, it has added AI features to automate parts of segmentation, experimentation, and message optimization, but the core product is event-driven messaging orchestration infrastructure. AI bear case: Braze is a feature-rich, engineer-forward customer engagement platform. As AI increases engineering-led companies' ability to build internal tooling, they may see Braze as a target for replacement, given that it is relatively expensive software that customers invest heavily in customizing. The rise of AI-powered mass personalization is an opportunity but also a threat for Braze, as that could lead to architectural shifts. Finally, AI may help adjacent players (Adobe, Salesforce) enhance their platforms, accelerating consolidation fears that have long plagued the business. AI bull case: Braze's capabilities are complex, omnichannel and hard to replace. AI offers opportunities for deeper personalization (aka value add for Braze), and running communications infra at hyperscale remains technically challenging and resistant to vibecoding. Recent results suggest that Braze is seeing strength in the business, perhaps from a wave of tech stack reconsideration since competitors Salesforce/Adobe are slower to adapt in an AI native world. Braze is led by a technical founder/CEO. AI traction: None disclosed, although the business saw improving NRR and organic growth in 2025. Adjacent AI-native startup summary: N/A- given it's enterprise/platform approach, there aren't any startups tackling Braze head-on or building on top of it today. Long-term, companies like Decagon/Sierra could leverage their CX positions to enter proactive communications and compete with Braze, though there is no evidence of this today. Management Quotes: "By applying state-of-the-art reinforcement learning and generative AI across an ever-evolving array of messaging channels and product interfaces, we help our customers leverage their first-party data to deliver more relevant experiences for their consumers and grow their businesses." "We are pleased to see customers using the full spectrum of Braze AI capabilities, including by crafting dynamic campaign content using the Braze Liquid Assistant, accelerating content production using Braze AI copy and image generation tools, ensuring strong clarity, impact and tone of messaging with Braze AI content quality assurance and delivering smarter product personalization with AI item recommendations." "Using Braze AI Decisioning Studio with reinforcement learning agents that independently experiment and identify optimal actions, they deliver deeper one-on-one personalization at incredible scale, managing approximately 5.1 quintillion permutations to select the optimal action for millions of their customers. The results generated a rapid and meaningful uplift in customer engagement, including a 12% uplift in app downloads and a 15% increase in conversion to premium memberships when compared to their prior strategy." "So when we think about the monetization of AI, and we've talked about this a little bit over the last couple of quarters as AI has just been introduced more generally from a monetization standpoint, we think about it in 2 buckets. So leaving aside Decisioning Studio, which obviously we are directly monetizing on a use case basis today. And then there's sort of 2 other flavors of AI that live in the tool. One is AI that is generally helping our users, our customers with the overall workflow and things that you invoke kind of once and then allow for kind of a broad-scale deployment of a particular canvas or campaign or content that doesn't really weigh on our own cost structure in the same way as things that invoke AI sort of on a repeated basis that are on a one at a time in real time, always on function. And so the things that are just kind of invoked occasionally for kind of large-scale deployment, sort of occasionally, that we would sort of include in the platform and largely not charge for those on an invitation basis. The things that are kind of operating one at a time in real time, we anticipate putting those into the credit framework and charging customers as they invoke the LLM usage, which, therefore, is going to have some impact on our cost structure over time. And so that's how we plan to incorporate that. We are not there yet. And so that is potential upside as we include that in the credits portfolio." Additional Commentary: Braze has similarities with many systems that VCs think will be relatively immune to vibe-coding- it operates a broad platform at scale, is heavily embedded, mission critical, low latency, etc. The key debate on the company has been its battle with Adobe/Salesforce for enterprises considering their ability to bundle- it is an open question whether AI helps those vendors compete or gives Braze room to further differentiate. The canary here would be if some of Braze's lighthouse customers opted to churn in favor of building/maintaining infrastructure like this themselves- it will be interesting to see if that happens in coming years.

@JaredSleeper Great stuff here! Thank you. For the Bull case, RL turns platforms into decision engines. The vendor that can safely run real-time experiments at scale is the default decision layer across channels, and that’s not easy to DIY.

As a bonus, here is who Braze gets compared to the most on G2 Crowd- though this sort of signal skews towards smaller customers and Braze generates >60% of ARR from customers with >$500k ACVs

We are so slow! Look at the Eglinton Cross town here in Toronto. That should be a national embarrassment. Connecting people to the value of going fast with quality is a must.

Best ads I have seen in a while @NordVPN ! #hatTip

@donnelly_b Cabbagetown!

If money wasn't a concern, which neighborhood in your city would live in and why?

@howardlindzon Those are delicious!

Caught a prison fish in the keys this am knows as a sheephead

Elections are for later. Vote Conservative, vote Liberal - that choice will come. Put down your partisan swords today and take a moment and listen to this speech and what is being framed. These times are not like any other.

Without comment!