Kushner & Iger’s $12.5B Lakers Deal, Grok 4.6 Launches, Anthropic’s Watermarking Move | Diet TBPN

Summary

TBPN Podcast Summary

Episode: Kushner & Iger’s $12.5B Lakers Deal, Grok 4.6 Launches, Anthropic’s Watermarking Move

Date: August 13, 2026
Hosts: John Coogan & Jordi Hays

---

Episode Overview

This condensed "Diet TBPN" episode focuses on three main topics:

• The unprecedented $12.5 billion sale of the Los Angeles Lakers to Josh Kushner and Bob Iger
• The launch and competitive implications of Grok 4.6, the latest AI model release from xAI
• Anthropic’s move to watermark all AI-generated content, and the regulatory and cultural impacts

Humorous tangents and hot takes pepper the discussion, including a satirical "Pothole Company of Los Angeles" pitch and lively debate over startup investor update etiquette.

---

Detailed Breakdown

1. The $12.5 Billion Lakers Deal

[00:02–09:46]

Key Facts & Details

• Headline: Josh Kushner (Thrive Capital) and Bob Iger (former Disney CEO) are purchasing the LA Lakers for a record-setting $12.5 billion (Jason, 01:48).
• Background: Mark Walter, who bought a controlling stake from the Buss family for $10B in 2025, is selling amid a federal investigation into private credit transactions associated with his firm (Jason, 02:51).
• The NBA Board of Governors’ approval is pending, but the hosts believe the deal is as good as done.

Notable Quotes & Reactions

• “With all of the technology disruption happening, it feels like these sports franchises could be more durable than many of the biggest companies in the world today.” – Jordy, [00:20]
• “Kushner and Iger made the offer to Mark Walter...officially became the Lakers majority owner last October after the NBA unanimously approved the deal.” – Jason, [01:48]
• Magic Johnson: “Lakers fans couldn’t have two better owners...He [Iger] will bring championships back to LA.” (Jason, paraphrased, [04:30])
• Luka Doncic: “Being a Laker means everything to me...no limit to the potential of this iconic franchise.” (Jason, paraphrased, [04:00])
• “Longevity sports is a portfolio diversifier. … Investing in sports properties as a hedge against higher risk investments.” – Jason, [08:41]

Financial and Strategic Insights

• Hosts note the durability of sports franchises as assets indifferent to technology disruption.
• Sports teams are increasingly seen as “hedges” or “portfolio diversifiers” for tech investors seeking stability amid AGI-driven change.

Humor and Tangents

• Extended riff on the state of LA's potholes evolves into a satirical business proposal – "The Pothole Company of Los Angeles." [05:36–08:33]
• “I might become the Batman of potholes. Sneaking out late at night.” – Jordy, [07:00]

---

2. The Grok 4.6 AI Model Launch

[10:14–14:39]

Key Points

• Grok 4.6 Performance: Delivers roughly the same results as Fable 5 Max at just 15% of the cost (Jason, [10:27]).
• Cost: 80% cheaper for input tokens, 88% cheaper for output tokens than competitor models (Jason, [10:27]).
• Grok 4.7 is teased as a forthcoming, significantly larger model with better performance due to inclusion of SpaceX data and pretraining updates.
• The AI model race ("model wars") is intensifying, with Nvidia jumping in via its NEMO Switchyard model router.

Industry Impact & Analysis

• Price wars between major AI model providers are accelerating (Jordy, [11:51]).
• Trust in benchmarks is low, with lived experiences diverging from advertised results (Jordy, [12:02]).
• SpaceX’s value has surged 40% in the past five days, possibly linked to AI/model momentum (Jordy/Jason, [12:16–12:37]).
• Grok is positioned as the “digital colleague,” with features tailored for work and high-value tasks (Jason, [12:51]).

Notable Moment

• “With each release, Grok is becoming a more capable digital colleague. The age of work is upon us. Grok for work. It’s happening.” – Jason, [12:51]

---

3. Anthropic’s Watermarking Move

[15:34–21:59]

Key Points

• Anthropic Announcement: All content generated by its Claude family of AI models will be watermarked (Jason, [15:39]).
  • Watermarks are not visually obvious; users must analyze text/files through a dedicated tool (Jordy/Jason, [16:05]).
  • Integrated as a regulation-driven measure, mainly influenced by the EU (Jason, [16:18]).
• Ben Thompson’s Analysis: The watermarking is implemented deeper at the model level, affecting all outputs globally to avoid geo-fencing complexities (Jason, [16:18]).
• Potential Issues: There’s concern about pervasive “AI bias” in crediting creative works if even minor edits trigger watermarks (Jason citing Ben Thompson, [17:30–19:46]).
• Technical Details: Watermarks work at the token level; only revised portions should be marked as AI-generated, not entire pieces of text (Tyler, [20:06–20:52]).
• Cat-and-Mouse: Tools to bypass watermarks (e.g., Pangram fighter) are already circulating, sparking an “arms race” between detection and evasion (Jason, [21:40]).

Notable Quotes

• “What the EU is doing is stealing the last thing humans have—creation—and demanding it be bundled with substantiation, effectively giving the AI the credit.” – (Ben Thompson via Jason, [17:46])
• “People think of a watermark as binary... it’s either going to say it’s AI or not. In fact, what it’s going to be...the peanut gallery online...it’ll be very binary for them.” – Jordy, [20:58]
• “How long till someone ships the Claude Watermark Remover app?” – Jordy, [21:35]

---

4. Startup Investor Update Etiquette Debate

[22:10–27:42]

Context

• Alfred Wallforce (Listen Labs) asserts: “Founders only send investor updates [when] winning. Nobody reads a bad update, helps. They just mark you as dying.” (Alfred, [22:11])
• The hosts push back: Regular updates build trust, even when things aren’t going well (Jordy, [22:34–23:51]).
  • Anecdote: An "assumed dead" angel investment resurrects with a $250 million exit—surprise, delight, and missed updates (Jordy, [23:18]).
  • It's harder to get bridge funding from silent investors (Jordy, [24:16–27:42]).
• “If you’re not updating somebody...it’s so much different to be getting those updates in real time, seeing how they’re dealing with adversity, seeing how they’re building momentum.” – Jordy, [26:54]

Memorable Exchange

• “Don’t do this. Just don’t do it.” – Jordy, on Alfred’s “only send updates when you’re winning” advice, [27:42]
• “It’s not one size fits all advice. I completely agree.” – Jason, [26:16]

---

Notable Quotes & Memorable Moments (with Timestamps)

• “With all the technology disruption happening, it just feels like these sports franchises could be more durable than many of the biggest companies.” – Jordy, [00:20]
• “Kushner and Iger...are being sold the Lakers for a record breaking price of $12.5 billion.” – Jason, [01:48]
• “I might become the Batman of potholes. Sneaking out late at night.” – Jordy, [07:00]
• “Grok 4.6’s performance is about the same as Fable 5 Max at an 85% discount.” – Jason, [10:27]
• “Price wars have definitely begun.” – Jordy, [11:51]
• “All of the Claude family models will be watermarked...the EU has achieved their dream of regulating the world.” – Jason, [16:18]
• “If you have AI edit [your work], it will be watermarked. People will say, ‘what was the prompt?’ even if there wasn’t one.” – Jason summarizing Ben Thompson, [17:30–19:46]
• “Don’t do this. Just don’t do it.” – Jordy, investor update advice, [27:42]

---

Timeline of Key Segments

| Time | Topic / Highlight | |---------------|--------------------------------------------------------------------------------------| | 00:02–09:46 | Lakers sale deep-dive; sports as a “durable” asset; investor/litigation context | | 05:36–08:33 | Tangent: “The Pothole Company of Los Angeles” satire | | 09:46–10:14 | Solar eclipse mention and AI stock superstitions | | 10:14–14:39 | Grok 4.6 launch details, model wars, Nvidia’s move | | 15:34–21:59 | Anthropic’s watermarking decision and regulatory implications | | 22:10–27:42 | Debate: Should founders only send investor updates when “winning”? |

---

Conclusion & Final Thoughts

• This episode delivers high-energy banter and informed takes on a blockbuster sports deal, evolving trends in AI modeling, and the nuances of regulatory decisions impacting creative work and startups.
• The Lakers sale is positioned as both a symbolic and strategic bet on everlasting brands in a world of relentless technological change.
• Grok’s aggressive pricing and feature advances signal a new phase in AI platform wars, while transparency and trust (in both AI outputs and startup communications) remain lively themes.

End Note:
"Have the best Wednesday ever." – Jason, [27:53]
“We love you.” – Jordy, [27:57]

Transcript

[Jason]

First we gotta talk about Josh Kushner, Bob Iger. They bought the LA Lakers. They're buying the lake.

[Tyler]

He's done it again.

[Jason]

He's done it again. They said he couldn't do it in

[Jordy]

the sauna this morning. I was thinking about the name Thrive Eternal. I think it's one of the best names for a fund ever.

[Jason]

Yeah.

[Jordy]

Like, Eternal is such a powerful statement and it fits the strategy so well. And it does. You know, people have been talking about this this morning, but it feels like with all of the technology disruption happening, it just feels like these sports franchises could be more durable than many of the biggest companies in the world today.

[Tyler]

Yeah.

[Jordy]

And I figured out how to put on a suit.

[Jason]

Yeah. Doesn't he look great? Look at this guy. It's incredible.

[Tyler]

I cleaned it up.

[Jason]

Yeah. I cleaned it up clean. Looking amazing. Yeah.

[Jordy]

Finally, truth be told, when we started the show, I went from. We started hitting the gym.

[Jason]

We did a little bulking. It was bulking.

[Jordy]

And I added. I added some weight and I basically over.

[Jason]

You added 45 pounds of lean muscle. You don't need to undersell it, Jordy. And you lost a little weight. Yeah.

[Jordy]

So I added a little weight and I basically grew out of my suit slowly. About two pounds a month.

[Jason]

Yeah.

[Jordy]

For 18 months.

[Jason]

It went exponential and it was a fast takeover.

[Jordy]

It became extremely uncomfortable to sit here even though all these. The suits. Suits are beautiful that we had gotten

[Jason]

made, but they got a little tight.

[Jordy]

I'm gonna have to save them for when I'm an elderly man.

[Jason]

Yeah, maybe.

[Jordy]

Because at some point we'll shed. We'll shed the muscle.

[Jason]

Cutting season's coming.

[Jordy]

We'll be lean, maxing. Summer's only cutting season is coming nine months away. 40 years. 40 years.

[Jason]

Yeah.

[Jordy]

But it's good to be back.

[Alfred]

Yeah.

[Jordy]

It's good to be back.

[Jason]

Yeah. Looks good. Let me take you through the story so everyone has the facts straight about this incredible deal for the Los Angeles Lakers. The headline is the Los Angeles Lakers are being sold to American businessmen Josh Kushner and Bob Iger for a record breaking price of $12.5 billion. Multiple sources have told this to ESPN. So Kushner and Iger made the offer to Mark Walter, who. Who only last year purchased a controlling interest in the Lakers from the Buss family at a then record valuation of roughly $10 billion. Walter, the CEO and chairman of holding company TWG Global, officially became the Lakers majority owner last October after the NBA unanimously approved the deal. The new sale will likewise require approval from the NBA's board of governors, which is scheduled to meet next month in New York. But looking at that picture, it seems like they've had this conversation with the relevant people. They're not some random businessmen coming in from out of town who and the

[Jordy]

seller is under federal investigation.

[Jason]

Yes. For some very complicated deals where some private credit transactions were put into an insurance firm that they own. Those affiliated transactions, if you originate the deal and then you put it in your insurance company's, on your insurance company's books, you need to, you need to disclose this now. You can actually do this. Berkshire Hathaway about 46% of their insurance assets are related party transactions. So there's a way to do it above board. The allegations are that he disclosed that only 3% of the insurance assets were related party affiliated transactions, but in fact it was much higher when you considered some of the other assets that had made their way through a capital allocator.

[Jordy]

Bit of a bad boy, even the bad boy thing.

[Jason]

Yeah, maybe. But you know, innocent until proven guilty. Obviously these are just allegations. He's just being investigated and a nice little return. Yeah. And he's denied all wrongdoing. But it does give you a little bit more context on what might be motivating a sale so soon after buying an asset that he was clearly optimistic about. So Kushner is the founder and CEO of venture firm Thrive Capital and co founder and vice chairman of Oscar Health. He also is the brother of Jared Kushner, President Donald Trump's son in law. Kushner currently owns minority stake in the Miami Heat, which he will have to sell to complete the Lakers deal, and was previously a minority owner of the Memphis Grizzlies. Iger, meanwhile, stepped down as Disney CEO earlier this year and has already made a major investment in Los Angeles sports. He and his wife Willow Bay, became controlling owners of the National Women's Soccer League's Angel City FC in 2024. The prospective new owners are already getting a warm reception from some of the biggest names associated with the franchise. Luka Doncic said, being a Laker means everything to me and I'm excited to get back to the court and bring a championship to la. I've gotten used to big changes over these last few years, but I know that no matter what, there's no limit to the potential of this iconic franchise. I look forward to meeting Josh and Bob so we can get to work building something special in LA together. You'll love to see it. Magic Johnson chimed in as well. He said Lakers fans couldn't have two Better owners. Johnson said, I've known Bob personally with for over 40 years. He has always loved the Lakers and basketball and he will bring championships back to la. So Magic Johnson is optimistic about this deal.

[Jordy]

Weirdly, this is making me optimistic on la. Overall, I've had a generally bad outlook. I have had extreme confidence that the weather is going to be very nice for a long time, but that's really been the only thing I'm willing to really lean on. But this is giving me a slightly more positive outlook.

[Jason]

And you've been pitching thrive on your new incubation. The Pothole Company of Los Angeles. The Pothole Filling Company of Los Angeles.

[Jordy]

Huge opportunity. It's really balance here for potholes.

[Jason]

Yes. Yes.

[Jordy]

Yeah.

[Jason]

Hiring the best AI research.

[Jordy]

No, I want somebody. I want somebody to do this. Figure out how to fix potholes quickly.

[Jason]

I think it starts with its own satellite constellation. You have to put up 10,000 satellites in low earth orbit with cameras pointed with.

[Jordy]

Actually no, small, small, effectively missiles fired at potholes.

[Jason]

Yes.

[Jordy]

And then a laser that sort of like melts the debris down to pull in the hole.

[Jason]

Yeah, it just sort of. You bring the molten asphalt up to space, it comes down, warms up as it travels through the atmosphere, lands directly pothole and smushes in, fills it in.

[Jordy]

But, but you know, people are always saying, oh, why is like you go to la, every other car is a G wagon. Well, you know why? Because the roads are absolutely terrible. You pretty much need something that's four wheel drive.

[Jason]

Yeah.

[Jordy]

Otherwise you're going to have a bad time. I was talking to. I was talking to a Maybach owner a couple nights ago and he had blown out multiple tires in the last year. I've blown out multiple tires last year. It's a disaster. Pothole. The Pothole Company of Los Angeles.

[Jason]

It's a banger idea, right? Yeah.

[Jordy]

I might become the Batman of potholes. Sneaking out late at night.

[Jason]

Yeah.

[Jordy]

Look, I could go out in this suit.

[Jason]

You could? Yeah. Like that guy who was cleaning up the FDNY corner call box. Like same thing.

[Jordy]

That's right.

[Jason]

Potholes in la. They're really, really bad. Every day I drive to work and I have to swerve into the other lane to avoid one on this one particular road that's just been a disaster for the last six months. Anyway, the last quote related to this deal, as lifelong NBA fans, we are deeply honored for the opportunity to become the stewards of the Los Angeles Lakers, one of the most iconic sports franchises in the world. Kushner and Iger said in a statement. They said they have immense respect for Jerry and Jenny Buss and intend to build on the family's legacy. I don't know what you're laughing at. Compete at the highest level and serve this extraordinary team, its fans and the city of Los Angeles. If approved, the deal would give the

[Jordy]

Lakers their third controlling ownership of potholes.

[Jason]

I think he got something.

[Jordy]

And Tyler says he's gonna film the missions. This would just be absolutely riveting content. You're out on the 4:05 2:00am, just you in a nasty pothole and a little bit of traffic and a little

[Jason]

bit of asphalt slathering it in there. Right. Is that what you're doing?

[Jordy]

Well, my big issue is when they fill in, they just do a really bad job. And so it basically creates a bunch of mini speed bumps and you get a sinkhole.

[Jason]

So then you just start. Sinkhole comes in, hand flow tamps.

[Jordy]

That's a whole other problem.

[Jason]

Yeah, that was crazy. Oh, whatever happened with that in LA was really bad for a while. I guess they fixed it.

[Jordy]

I would like to get somebody to throw on a tinfoil hat and figure out why a bunch of pipes started exploding everywhere around the city.

[Jason]

Somebody's gotta ask the question. Sag Harbor Capital says while there are plenty of issues you can point out with sports investing, I think this actually makes a simple yet incredibly compelling case. Longevity sports is a portfolio diversifier. Investment firms and wealthy individuals now view investing in sports properties as a hedge against higher risk investments. Very interesting. Everyone has always said sports investing is just, it's a trophy asset. It's just because you like the team, but you're locking up a lot of capital. And when you're long AGI the future, the most cutting edge technologies. Also building a portfolio with some things that are going to be around forever, no matter how many robots there are, makes a lot of sense. And so that that balancing effect seems to be a new thing that's driving demand in in sports investing. Do you agree with that?

[Jordy]

I do. Fits the strategy, yeah.

[Jason]

Spoor agrees. If you truly believe in AGI, buying a massive sports franchise is actually one of the best financial decisions you can make right now. And we've seen hints and rumors of this for a while. Anything else on the Lakers? We're going to talk to Darren in just 13 minutes about it. Get him to explain.

[Jordy]

He's going to walk us through the whole rule set. Yeah, we're very excited about that. Anyway, apparently there's a solar eclipse going on right now.

[Jason]

Oh, is that going on Right now I saw a link to a whole bunch of different webcams that you can use to watch it. The last time a couple of years ago, there was a full solar eclipse, not a partial one. And everyone who went said it was one of the most insane experiences of their life. So I would highly recommend it. Tyler?

[Tyler]

Yes, it's in Europe right now. Full solar eclipse.

[Jason]

Full solar eclipse, Yep.

[Tyler]

Wow.

[Jason]

Should have gone. I missed this one. When's the next one?

[Jordy]

What does it mean? What's the meaning? Tyler, what does it mean?

[Jason]

It's a great omen for high beta AI stocks. That's what I choose to read into it. Lever up. That's what the. That's what the eclipse is trying to tell you. Grok 4.6 is roughly the same performance as Fable 5 Max at an 85% discount. 80% cheaper for input tokens and 88% cheaper for output tokens. Pareto dominance as Gavin Baker. Grok 4.7 will be significantly better, as is a much larger model with the cursor and SpaceX data included in pre training. So the headline is SpaceX sort of jumped to the frontier, at least on these benchmarks. We'll see how this comes out, how people adopted. Obviously Cursor has huge, huge usage all across the enterprise, across businesses. Grok has been slower on adoption with a bunch of different pushback there. But the model seems to be catching up. The teams seem to be gelling and things seem to be pretty good over there. We'll see how. I mean, I'm sure on X we'll see lots of people demoing examples of what they build. With Grok 4.6 there's already some games that have hit the timeline. There will be a whole, whole bunch more on benchmarks, so.

[Jordy]

So seems incredibly impressive. Excited to get reactions and get people's lived experiences with the model. Yeah, Price wars have definitely begun.

[Jason]

Yeah.

[Jordy]

You know, we saw this with Metta, we're seeing this with Grok now. And then third, I would say trust in benchmarks has to be at an all time low.

[Jason]

Yeah.

[Jordy]

Right. There's been enough moments, releases so far where the benchmarks looked amazing, but then the actual lived experiences is, you know, less, less significant. They also had a great launch yesterday, so they're building momentum.

[Alfred]

Sure.

[Jason]

SpaceX is up almost 40% just in the last five days. That's a pretty.

[Tyler]

Yeah.

[Jordy]

And I'm so curious. I wish, I wish you could figure out a way to parse out what, what's driving that. Is it the semi analysis? Yeah, Thesis, semi Analysis.

[Jason]

That felt like Friday.

[Jordy]

Friday? Well, yeah, that was Friday at the close.

[Jason]

It popped during the day though. Right, so.

[Jordy]

But then up 10% today.

[Jason]

Yeah, I don't know, maybe it was leaking out.

[Jordy]

CNBC says Space X short sellers are running out of bullets as stock rebounds more than 30% off. Running out of ammo. We're running out of ammo.

[Jason]

I'm tired, boss. I can't keep buying these prices. You gotta buy more. Excited to release Grok 4.6. With each release, Grok is becoming a more capable digital colleague. The age of work is upon us. Grok for work. It's happening. 4.6 is significantly better at difficult tasks and knowledge work. It combines opus, class, intelligence and polish with very low cost and high speed, which people love, even just in the cursor workflow. The more iterative approach instead of having to come back a day later and see what goes on, there's a lot of value to that workflow. In particular, Grok 4.6 has a very impressive test time curve on Cursorbench. In a league of its own, they did more mid pre training on the Grok 4.5 checkpoint and newer SFT stages with Grok 4.5 traces and model based filtering. Once again shows how important a good SFT checkpoint is. Interesting. Getting some more details on what's going on with Grok. They also launched Grok Bot. We recently hired a new teammate. Bots are AI teammates that do real work for you and they sign into your tools, use them just like you do and come back with finished work. So I guess Gwark isn't going to happen. It's going to be Grokbot, which actually it has a nice little rhythm to it. I like Grokbot also. Very good results on the Artificial Intelligence index. Joining the frontier at 61. Strong agentic performance, lower cost obviously. And the model wars are upon us. Nvidia is also joining the model wars, selling weapons to both sides, potentially. Nemotron 3.5 Lightning, their open source model came out, but they also launched a model router, Nemo Switchyard to deliver faster, smarter, more efficient agentic switchyard. Switchyard's a good name.

[Jordy]

Good name. I like that kind of railroad connotations, which.

[Jason]

Oh yeah, which maybe that's interesting. Yeah, maybe Gotta go through Ben Thompson stuff. It's some point, but interesting to see where this goes. I mean it does feel like if you're a organization that has Nvidia chips, you have your own data centers because you're a mid scale, maybe not a hyperscaler but you're pretty big. Being all in on the Nvidia stack and then using their model router could make a lot of sense. Like it's a logical place for it to live. And then it has more integration with the rest of your Nvidia stack because you have a whole bunch of GPUs racked for the lowest possible cost. If you're cost optimizing, you'll probably eventually have your own racks and your own Nvidia work data center, basically. Anyway. Should we debate anthropics watermarking?

[Jordy]

Yeah, dive into this. I don't know that I have a strong take yet.

[Jason]

Ben Thompson has a strong take on anthropics watermarking. So the basic news is that you might soon be able to know for sure if something came from a human being or from Claude. Anthropic said this week that text and files generated by its family of AI models will be watermarked to let consumers know now they won't be watermarked. Like, you'll still have to take the text and then go to their watermarking tool and validate that it was. That it was.

[Jordy]

Bill Gurley was making the point that traditional watermark on an image is very visible.

[Jason]

Yes.

[Jordy]

Just like immediately visible.

[Jason]

Yes.

[Jordy]

And so it's not actually. It's more like basically we're going to leave evidence that was generated. Breadcrumbs.

[Jason]

Yeah, breadcrumbs. And so in theory, a LinkedIn or, or an X could integrate with an Anthropic API to check the watermarks. But the interesting read on this is that this was actually implemented by the regulation in the European Union. So anthropic in theory could watermark. Could only watermark text in Europe, but the EU has achieved their dream of regulating the world, according to Ben Thompson. And so they actually, instead, I think the idea is instead of, instead of doing a last step at the end of the generation to add the watermark in, it's baked much deeper into the model level, which means it will be watermarked whether you're in Europe or not. And so Ben Thompson explains a little bit on how it works. Obviously they're not explaining too much because then you can avoid it easier.

[Jordy]

Yeah, it's so, it's so interesting, like how much of their revenue is coming from Europe, given how compute constrained they are. Like, was there. I wonder if they ever considered like, hey, let's just say if, hey, if you're going to, if you're going to go forward with something like this. Yes, we're just going to exit the market entirely.

[Jason]

But I don't think people care. I don't think Anthropic cares about watermarking. I think they're fine, I think it's nice, I think it's a good thing. And then also I think business users don't care because they're like, yeah, I want AI generated code all over my code base. I want it to be clean and efficient and I want it to just do the job and I'm totally fine with it. That's not my business. And so I don't think people will be upset about this, at least in the enterprise. Ben Thompson points out that he is somewhat frustrated by this because even if you're using an LLM for proofreading and making individual changes in your text editor, like you just upload your blog post to Claude now and you say, oh, like, what could be better about this? What's wrong with this? Did I make any grammatical errors? Did I make any spelling errors? If you go and implement those, there's a small risk. It seems like a very unlikely risk, but there's a small risk that even in just accepting those changes, you accidentally insert a little bit of the watermark in, or potentially all of the watermark in, and then you don't get credit potentially, because people will be like, oh, this is AI. When really it was like the last step. That was AI. We had a back and forth on this a little bit where I would write an essay by hand, we would throw it into ChatGPT to proofread it and it would put in a bunch of EM dashes. In the earlier model, it was really heavy on EM dashes. And so people would just be like, oh, this is AI. And it was like, well, like the last step. But it's like, but does that matter? It does matter to a lot of people. They do care about that. But it's odd. And Ben Thompson's point is saying what the EU is doing is stealing the last thing humans have creation and demanding it be bundled with substantiation, effectively giving the AI the credit. Even if you come up with a great idea, you write a great book, if you have AI edit it, if it has a watermarking feature, it will be watermarked and people will be like, nah, yeah, it was all, what was the prompt? And it's like, well, maybe there wasn't a prompt. Maybe there was a hole process that you went through to develop your idea and then you took it at the last step and so that's the source of frustration. At the same time, it does feel like there's ways to work around this. If you become more collaborative with the AI system, you can avoid the watermark by only using it at one certain step and then doing an editing pass yourself. People have been going back and forth these all. All day long. He says the human that gave AI the text to proofread or even the prompt to generate writing doesn't matter to the bureaucrats in Europe. Everyone is worried about being replaced by AI. The EU is mandating it. That's what he says. He's got some harsh words. What do you think, Tyler?

[Tyler]

Yeah, I mean, so I don't think the actual, like, underlying the way they watermark it is, like, public, but presumably it's like, basically during the sampling process when you're, like, generating tokens.

[Jason]

Yep.

[Tyler]

Right. It generates some distribution over the tokens, like, which one is it going to select? And then it uses, like, a slightly different method than just choosing, like, the most likely one.

[Jason]

Yep.

[Tyler]

And that way you can, like, see, oh, this is, like, less likely. So it's probably, you know, written by AI. But, like, if you basically have, you know, some essay that you wrote and you replace, like, one sentence, then, like, 99% of the tokens are still the same. Then, like, it's like, it's. It shouldn't mark the whole thing as being, like, AI written. That doesn't make any sense. Like, only the sentence that you changed would be marked as AI written.

[Jason]

Oh, sure, sure, sure.

[Tyler]

Or whatever.

[Jason]

Yeah, yeah, yeah.

[Tyler]

It doesn't, like, say the whole document is air.

[Jason]

Yeah, yeah.

[Tyler]

Like this sentence.

[Jason]

No, no, no.

[Tyler]

These tokens are very.

[Jason]

People think of a watermark as binary. It's either going to say it's AI or not.

[Jordy]

In fact, what it's going to be. I think people are going to just look, you know, like the peanut gallery online is going to be. It'll be very binary for them.

[Jason]

Sure.

[Jordy]

Like, if somebody puts out even an Instagram caption, they'll get dunked on, even if it's one sentence. Right.

[Tyler]

Like, if you look at the Google Synth id, so Gemini has done this for, like, a few years. If you look at their blog that, like, explains how it works, it will, like, highlight specific, basically words in the blog. In the blog or whatever.

[Jason]

Sure.

[Tyler]

It's like, okay, this is probably. It's not saying the whole thing is like, 01.

[Alfred]

Yeah.

[Jason]

So, yeah, yeah.

[Tyler]

I think. I think this is kind of like blowing it out of proportion.

[Jason]

Yeah. I don't. I don't think that many People will have a problem with it.

[Jordy]

How long till someone ships the Claude Watermark remover app?

[Jason]

I mean, somebody shipped a Pangram fighter, like, yesterday. Something where you take the text, you put it in this model, and then it. It reads 100% human to pangram because it rewrites it in a style that Pangram can't detect. And so now the Pangram team has to go figure out how to defeat that. It's all a game of cat and mouse.

[Jordy]

Let's head over to the timeline. Alfred Wallforce over at Listen Labs. He's been on the show before, coming in with one of the hottest takes.

[Jason]

Yeah, Got some pushback.

[Jordy]

Said founders only send investor updates.

[Jason]

Yeah, give me the boat horn when you're winning.

[Alfred]

This has sent three in its history. All fundraise announcements. Nobody reads a bad update helps. They just mark you as dying.

[Jason]

I think your crazy voice sounds crazier than mine.

[Alfred]

Let me say. Let me check. Oh, it sounds about the same. It sounds about the same.

[Jordy]

But, yeah, this is just. I think this is terrible advice. Except if you're winning, I think it should be founders if you're winning, only send investor updates.

[Jason]

Yes, I do.

[Jordy]

It's just like. It's just such bad advice because I've had so many. I mean, I've had instances, like kind of surprise and delight moments where I haven't been getting an update from a company in a year. And I just assume the company is, like, dead and dying, and then they pop up with an update. I had this happen where a company that I thought was fully dead. I didn't even want to. I almost felt bad asking. It was a small angel check. I thought they were dead until I got an update that I needed to sign something because they were selling for like, wow, quarter of a billion. Updates are how you build trust. It's how people start to process. Like, how is this team and this founder performing even if they're not winning yet, Are they making progress? Are they iterating quickly? It's how. And if you're not sending any updates and then suddenly you need to raise more money, like, you need a bridge round or something like that, the likelihood of getting that round done is much lower because you just don't have the. You don't have the tr. So people can't get a good read on. Like, okay, do I have more conviction in this team today than when I invested? Even though they haven't figured it out yet.

[Jason]

Okay.

[Jordy]

Right.

[Jason]

So I'm taking the other side of this with some reasonableness. Maybe I don't know. You can debate me on this. So do you think it's important to send investor updates when you're a pre seed company? Like purely in the idea maze, very early stage, the revenue is zero every month. You're just sort of like, you know, out figuring out the first product.

[Jordy]

I mean, at that, at that stage, it depends. Hopefully you only have like one or two investors. Yeah. But it's still.

[Jason]

So you're maybe just doing.

[Jordy]

It's still important to see like, it's valuable to see a snapshot. It doesn't have to be at that stage. If I get an investor update, that's paragraphs and paragraphs and paragraphs, I'm like super, super bearish.

[Jason]

Yeah, yeah, totally.

[Jordy]

But it's still better to just have like even quick bullet points. Hey, here's what we're working on. Yeah, we'll come. We'll. We'll have another update in a month.

[Jason]

Okay, so, because listen, Labs is only a three year old company and they raised their Series A in April of 2025. So. And he, and there's a, there's a rumored Series C going on right now. And so if he sent three investor updates, that might be April of 2025, January of 2026 and August of 2026, which is like every six months. Which doesn't feel like that crazy of a cadence. It's like, yeah, it, it's not monthly, but it's a pretty regular pace because they're raising money regularly. So it's very different. If it's like, okay, this is like a more like less like hot time in the market. You're not going to be fundraising every three months. You're going to be raising money every 18 months. Realistically, every 24 months. It's a harder industry. And so as an investor with a series A check, I want to know, okay, how much Runway is actually left because. But that's not the equation here. So little bit of a unique. This post reveals a little bit more about the level of fundraising activity that you can only send investor updates and still be emailing all your investors every six months.

[Jordy]

Yeah.

[Jason]

So there's a little bit of that.

[Jordy]

And a lot of this comes down to Alfred. Listen, Labs has been crushing. I remember when they first came on the show, it seemed like the company made a lot of sense. They, they quickly raised follow on rounds from great investors. I'm sure they're still doing well. You wouldn't post this if you weren't winning. Still sure they're doing well. But like, just really, really bad. Just really, really bad advice.

[Jason]

It's not one size fits all advice. I completely agree. The other.

[Jordy]

Yeah, and the other, the other thing I would say is like, how do you know a company that's building in public is struggling? They stop building in public. No, they haven't. Like, I've seen this like at least 10 times right now. You see the company that, that's got all this momentum. Founders, like, I'm sharing our.

[Jason]

Here's our best because they're just proud of it.

[Jordy]

And they usually will string together like six months of that.

[Jason]

Yep.

[Jordy]

And then they go dark. It's like, oh, what, what? How you don't want to build in public when you're not growing anymore and revenue is declining. Or, or you were, you were a thin wrapper. Whatever it is.

[Jason]

Yeah.

[Jordy]

If you're not updating somebody. Like, I, I had one of the 70 odd angel investments that I've made. Hadn't heard from them in a year and then they texted and said, hey, are you free to catch up? And the ask was basically like, will you do a bridge round? And it's like, yeah, I can ask all the questions to try to understand what has happened between now and when the investment was made, but it's so much different to be getting like those updates in real time, seeing how they're dealing with adversity, seeing how, how they're building moment momentum as a company and. Yeah. So anyways, don't do this. Just don't do it.

[Jason]

Don't do it in every case.

[Jordy]

Don't listen to Alfred, the founder.

[Jason]

I'm sure it applies to some companies.

[Alfred]

Companies.

[Jason]

I'm sure it applies to some companies.

[Jordy]

It's been an honor and a privilege

[Jason]

and have the best Wednesday ever.

[Jordy]

Please do.

[Jason]

Goodbye.

[Jordy]

We love you.