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 13 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
Junestrand: Fine-tuning foundation models will not be a difference maker today
“It didn't do that back in 2023 as a starting strategy. I doubt that that's going to be the difference maker today.”
Junestrand: Anthropic's Opus 4.5 is effectively AGI
“I think it's gotten to the point now where I think you should just coin it AGI and focus on optimizing the cost pretty much basically.”
Junestrand: Frontier AI model providers are already completely commoditized
“I think you're right in that the model providers are not the winners, because that's completely- Commoditized.”
Junestrand: Harvey made fundamental product errors focusing on early fine-tuning
“I think they made a lot of fundamental errors around, for instance, fine tuning and spending time on the wrong things.”
Junestrand: Klarna and Spotify product leaders lack incentives to be curious
“There's so many product directors at Klarna or at Spotify or these companies that have made it, that they're not incentivized to be curious. In a way, they're kind of incentivized to keep the company on the track it is.”
Junestrand: Legora built proprietary agent architecture because LangChain was too bad
“Basically Langchain was too bad at the time, so we built our own agent architecture, and that, we did that very early, which I'm very proud of, and that was like the right direction to continue on.”
Junestrand: Building an AI-native law firm is not a winning strategy
“Don't think that owning the entire service layer and software layer is a winning strategy. Basically building an AI native law firm.”
Junestrand: Early-stage founders should trade higher dilution for top-tier VCs
“No, because I don't think I would have gotten benchmark at a 10% dilution. And I think the VC that you get in the beginning, at least as a first time founder, is quite important because you don't want to get people in your corner who are going to get you off t…”
Junestrand: Legal AI software products currently lack stickiness and are easily replaced
“And I frankly think that so far, none of the products are super sticky. They're quite easy to interchange, which has been very effective for us because we've done a lot of rip and replace.”
Junestrand: Opus 4.5 Performs Like a VP, Not an Unintelligent Employee
“Managing an employee who wasn't very intelligent, whereas Opus 4.5 feels like a VP. You give it, here's the thing I want, go execute, and it just does it, and it's amazing.”
Junestrand: Grok Will Not Be a Top Model for Enterprise Legal AI
“No, we're not going to throw Grok in there at all.”