The Ledger
Every statement that passed quotation and attribution checks. Mix any filter with any other: certainty 1/5, debate potential 5/5, or both at once.
why aren't all 2,724 resolved? a statement only gets an assessment when the public record can support or contradict it. opinions and what-ifs never can, and 0 checkable ones are still open, waiting for their date. predictions held up or didn't; assertions are supported or contradicted. on every card: ▮▮▮▮▮ certainty · ▮▮▮▮▮ debate potential. speakers are clickable
O'Driscoll: Startups Cannot Sustain Both High Compute Costs and Sales Teams
“You either have compute that sells itself or people who sell it, but you can't do both.”
Marketers Can Produce Minutes of Commercial Video Daily Using AI
“Essentially, what happens is that essentially what we're seeing here is that content that's like one social media marketer can make several minutes of commercially viable videos, social media videos a day.”
Lemkin: Managing 21 AI agents requires more cognitive load than managing humans
“It is more tiring to be managing 21 agents. It is not. I don't know that it's more work. I'm still reflecting on it, but it is a higher cognitive load than people.”
Lemkin: AI Agent Goal Should Exceed Human Performance, Not Replicate 80%
“And so your goal should not be to reproduce 80% of your best person, but the goal should be to find areas where the goal is higher than human.”
Tavakoli: Text-to-SQL fails in BI without deep semantic data understanding
“I think the hard part for the standalone BI tools was they historically really had no semantic understanding of the data, right?
You didn't extract of a data warehouse.
They stood there.
And when they tried to add talk to data, it was mainly just, let me conve…”
Lemkin: AI support, sales, and marketing functions are converging into one agent
“If you talk to folks like at Finn or Intercom or Gorgias or Sierra and others, they're all saying that there's convergence that support sales and marketing are all converging because once you can AI-ify them, you don't want to have three different agents, righ…”
Lemkin: AI Autonomy Requires Humans on the Loop, Not in the Loop
“To get to the level of autonomy you want, you don't need humans in the loop. Okay. That that's not autonomy, right? What you need is humans on the loop constantly, the exceptions kicked out to them.”
Lemkin: High-ROI AI Agents Can Command 2x to 10x Software Pricing
“That will, in many cases, either deliver instant ROI or really replace humans, and it is so jaw-droppingly valuable that you can charge two to 10 times more than your software.”
Masad: AI context windows now exceed human context capacity
“I think we've already crossed a threshold. Yeah. For which the context links Is larger than any context that human can have. Right. And I think that's huge because, I mean, just think back two years ago we had 16 K context link. Right now we have a one million…”
Masad: AI agents work better with file systems than SQL
“Agents really like file systems. They like them better than SQL. They're really trained on doing grep and doing kind of search in the Unix command line.”
AI Limits Feature Differentiation Moats to Just One to Six Months
“So feature differentiation is really hard to maintain as a mode. You might be ahead for a month to six months, but then everybody will catch up because AI is writing so much of our code.”
App Building Is Commoditized; Differentiation Has Shifted to Value Creation
“We really believe that building something is already becoming a commodity.
Everybody and their mother are going into competition of creating an app builders.
How do you create value out of what you've built? Whether it's your personal value for your company or…”
AI Companies Nearing $500M Revenue Must Recapture Product-Market Fit Monthly
“Even Approaching half a billion dollars in revenue, we're still on the product market fit treadmill, because category is evolving so fast, both from technological perspective, from customer perception, that we still need to keep pushing the product at the maxi…”
AI Builders Create Internal Tools in Hours Instead of Buying Software
“So often you'd go and try to do RFP and buying software and I was like, oh, let me just go build it in an hour because let me see how far I can get there. So we constantly pressure testing our own boundaries of how much of our lovable we can build on lovable o…”
Lemkin: Buy off-the-shelf software rather than building custom AI agents
“The biggest takeaway is don't, still don't build it if you can buy it. We love Tenk and QB, they're part of our families, and I hate to say this, but we would dump both of them if there was an off the shelf tool. We wouldn't turn them off. This is the strange …”
Lemkin: Running AI agents permanently in dev environments preserves infinite context
“If you run them in the dev environment permanently, like we're doing, you get an extra agent that has an infinite context window and infinite history.”
Lemkin: Hire a VP of AI instead of running change management
“What's better is if your team is Lagging or resistant, just go hire one reasonably senior person that is all in here. One. And folks will start quitting when they don't want to work at that pace anyway. You got to hire them ideally at the VP level above, and t…”
Lemkin: Nobody will pay for '60% solutions' in AI era
“If you build a 60% solution. No, one's going to pay for it.”
Lemkin: Startups must view FDEs as retention drivers, not cost centers
“View FDs as one of your most important assets to getting renewals, retention, NPS, word of mouth, virality. Don't look at FDs as a cost center. This was the mistake we made in customer success a generation ago. And this is why customer success was cut to zero …”
Lemkin: Agentic API quality will drive vendor switches over human sales outreach
“It might be the API that gets us to switch, not the humans, not the chocolates, not the gift certificate, but if the API is better, works better with our agents, we might spend a couple hours. It takes to move over our key accounts to ramp because it's the age…”
Lemkin: AI SDRs fail unless an outbound sales playbook is proven first
“If you haven't gotten any outbound to work, Buying an AI to do it isn't, it's not going to help you. You have to prove it. Any of the agents we use, any of the dozens of agents we use, we are taking a playbook that at least works a little bit.”
Ibarra: Every new AI agent requires a two-week blackout period
“Even with the help we get, like, I feel like every new agent costs us at least two weeks of what I call either a blackout period and somewhat chaotic of, like, I kind of have to ignore the other agents anytime I bring on a new agent”
Ibarra: AI agents idle and waste money without daily human management
“You can't, you, it just can't afford to wait a whole week, or like you're, or you're probably just wasting, honestly, you're probably wasting your money if you don't spend the time to do it daily, because a lot of these are waiting for you to give it inputs, r…”
Ibarra: Managing AI agents degrades your patience for managing human teams
“I do feel like now that I've flipped and switched to managing more agents, I think I've probably become slightly worse in managing the people that are, you know, that work with SaaS or especially leading up to SaaS Radial.”
Lemkin: Vibe Coding Apps to Save SaaS Fees Makes No Economic Sense
“So there's no economic justification for software costs to building these apps, right? You can buy them for 10 grand, like it makes no economic sense. So you have to really need something that you can't buy today.”
Lemkin: Software companies without accelerating growth are not AI companies
“If growth isn't accelerating, you're not an AI company.”
Lemkin: Vibe coding cannot build Salesforce, only demo-day prototypes
“You're not going to VibeCode Salesforce for real, but you know what you can VibeCode? Something for demo day that looks really good.”
Lemkin: VCs now expect startups to hit $100M ARR in one year
“Investors are expecting insane levels of growth. Insane levels of growth, right? They want, like, the idea that you can go from one to a hundred in a year is now seen as what you want to invest in.”
Lemkin: Great compounding startups are unfundable today versus two years ago
“There's no great answers to the question that some great businesses will compound to Epic rates, but they're unfundable now, and they would have been fundable two years ago.”
Lemkin: Software platforms must own native agents or lose all value
“You have to own the agents on your platform, or the agents will take away all the value.”
Using LLMs for Marketing Defeats the Purpose by Reverting to the Mean
“In marketing, for example, marketing is all about standing out. You know what large language models are all about? Not standing out. They're all about reverting to the mean, actually. And that's literally what they're synthesizing a lot of information and givi…”
Lemkin: Deeply trained AI agents cannot succeed through self-service setup
“A lot of the agents we use are pushing down market to be more self-service. So far that doesn't work. So far, I will say for the most part, agents that require deep training cannot be self-trained.”
Adams: Employees needing constant convincing during AI pivots become liabilities
“Culture runs really deep. So you have to be wary of the people you need to keep convincing because they're going to be a liability.”
Adams: In AI products, building UI is easy while infrastructure is hard
“AI products don't look like this at all. It's actually the inverse. Building the UI is the easy part. That's the small part. The hard part is building great AI infrastructure, building products that actually truly work at scale.”
Adams: AI Product Marketing Must Focus on Infrastructure, Not Features
“A lot of AI products do look the same. A lot of them are now AI agents. They all kind of look like some kind of chat interface. They all do different things, but they all kind of look the same. It's hard to tell them apart. Look at five competitors. They all k…”
Lemkin: 99.8% of companies should use human managers over custom AI orchestration
“I don't think that is the right answer for 99.8% of the world. It might work for owner, and I'd love to learn from you. It might work for us, but most folks should not think about, most folks should realize you need a nerdy human managing your agents.”
Lemkin: The 20-30 Hour Work-From-Home Middle Ground in Sales Is Gone
“You could have it all. You could have lifestyle, quality of life, work from home, work for 20 or 30 hours a week, exceed quota for a lot of structural reasons. That doesn't exist anywhere today. Even the fastest growing startups they're either lean or they're …”
Anderson: SaaS systems of record will beat AI startups at parity
“If you were up against a competitor who's an AI only product and you have the system of record, I promise you, your customers do not want to leave. They are far more interested in staying right where they are and having their data and their documents and all t…”
Anderson: Incumbent SaaS can underprice AI startups using blended gross margins
“Your AI only competitors, they struggle with margins badly. They have a really hard time with margins because they have all their LOM costs are super high. So you get to do something kind of savage. You get to sell the AI products at a lower price point than y…”
Lemkin: Vendor counts are flat or shrinking outside of AI and incumbents
“If you're not in existing invest vendors that are able to increase pricing and you're not in a top AI initiative, the number of vendors are shrinking or flat.”
Lemkin: VCs Abandon Older Startups to Chase Early-Stage AI Hits
“If you've been doing this for a while, your VCs are just going to check out on you. They're just, it's just not again, like putting all your money into anthropic. It's just natural. If your company is around six, seven, eight, nine, 10 years, and you're lookin…”
Lemkin: AI Must Replace Humans to Secure Enterprise Budgets
“AI is about more than being productive. Ultimately it's about replacing or augmenting humans and delivering relatively rapid and massive ROI, much more than traditional workflow and database software.”
Lemkin: SaaS Profitability is Meaningless Without Zoom-Level 50% Margins
“Profitable is not enough unless you're utterly insanely profitable like Zoom. For every dollar Zoom brings in, 50 cents goes to the bottom line.”
Lemkin: AI Expands TAM by Enabling Cursor to Charge Thousands Per Seat
“Atlassian got incredibly successful selling JIRA and other products for five bucks a month with discounts to folks at companies. Now folks will spend 400 dollars, in some cases, 5000 dollars on cursor. That's why it goes to a billion.”
Lemkin: AI SDR startups succeed because replacing human labor expands pricing and TAM
“These guys are blowing up because the products unlock huge TAM expansion by replacing a lot of humans, a lot of SDR. And none of these guys would be remotely as successful if they had to charge like traditional SaaS companies.”
Lemkin: Growing 50% is a Loss if an Equal Competitor Grows 100%
“Even if you grew a 50% this year, if your competitor grew a hundred percent and you know, you're at the same AR, sorry, got a normal idea. You lost this year.”