Surprise AMA 08/22/2025

Hi, this is Charles Hoskinson broadcasting live from warm, sunny Colorado. Always warm, always sunny, sometimes Colorado. Today is August 22nd, 2025, and it is a Friday. I am just about to go and do this crazy thing with the Monroe Institute and enjoy all of the magic and amazing things that have happened this year. I just got back from Jackson, Wyoming, where I was at Salt hanging out with AJ Scaramucci, Anthony Scaramucci, and all the other people.

I did a round table with the Federal Reserve, and it blew my mind. In 2011, I was one of those "audit the Fed" guys running around saying we needed an audit. Now I'm sitting at a round table with Mickey and Governor Waller, and it was a great discussion. Sergey Nazerov was there, Anatoli from Solana, Ilia from Near, and Brad from XRP. Some nice people, and it was good to see some old friends and new friends, and make some deals here and there.

Everything is coming along. Midnight is progressing really well, and Cardano is also moving forward. We're pushing ahead. Redemptions are looking strong—over 70,000, give or take about 1.6 billion worth of tokens redeemed so far.

We just keep pushing forward. The 60 days plus the scavenger hunt and the remorseful redemption pool is a really nice layered way of distributing a token and getting a lot of other ecosystems involved. In fact, I had a great conversation with Brad about XRP, and it would be a lot of fun to do something there. Lace is going to support XRP before the end of the year. I'd love to do a joint event with some of the people on that side of the aisle and see if we can get John Deaton there, among a few others.

Overall, the Clarity Act is looking good. They’re going to have a whole write-up probably by the end of September, and a lot of cool stuff to do there. Hopefully, with any luck, we can get it passed in October. Genius is in the market. The Fed doesn’t know how to regulate Genius.

They were given a task of regulating without any statutory authority because there are three agencies: the Federal Reserve, the OCC, and the FDIC. They have to somehow make rules, even though only the OCC apparently has anything to say about it. That’s going to be tough. Maybe we can handle those things differently. But all in all, I'm pretty happy.

A lot of things are closing out, and new things are beginning. There’s a lot more to say and do with Cardano at Token 2049 and Midnight at Token 2049. We're also getting ready for the Asia tour, going to be in Vietnam, Korea, Japan, Singapore, and going all around. IO is coming to Asia. You guys have to see the vending machine; I think everybody really loved the vending machine at Rare Evo.

We have a whole bunch of stuff there. Now, let’s go ahead and get back to it. The first question: I would love to hear you address the decision to limit the CF from claiming Knight tokens. The same rule applied to IG and Ego Genesis distributions. Well, we built it.

It’s my money; we can do whatever we want to do. There was a disclaimer that said it would undo burden and harm to the network. First off, it’s not clear who owns the CF ADA. It’s community property, but it’s under the control of the Swiss government. I’ve always taken the position that the Swiss government stole the ADA.

The intention of that founding entity was for it to be a members-based organization, and the board to be elected and controlled by the community. When the Swiss government came in and took over the board, basically the assets were permanently entombed in Switzerland. When you have addresses that are either custodial or unclear, you can blacklist those addresses. What would happen? They’d come in and instantly be adversarial, abstaining or voting against anything the Midnight Foundation is doing and asserting that they have some sort of governance control.

Let somebody else have that Knight who’s going to do something more productive with it. They had every opportunity to try to come into the fold. We spent $23 million so far on helping CNTs out in the ecosystem. There’s this lie that gets pushed again and again that Midnight is separate from Cardano and does nothing for Cardano and it’s a different ecosystem. Well, we just announced today the Copper partnership.

What does that mean? It means that any exchange that uses Copper as a custodian can now support that asset. It’s the same for all the other custodians that are coming in and the relationships like Bitcoin.com, blockchain.com, and Brave.

They didn’t just agree to support Midnight; they agreed to support Cardano and Cardano native tokens, which Midnight is one of. There’s been zero support from the CF side. They have this gigantic endowment, and they should be entitled to go and receive a bunch of Knight tokens to do that. Also, members of the organization are low-key soft accusing us of stealing money. Members of that organization say that it was self-inflicted.

All this stuff about ADA redemption—were there any apologies? Was there any attempt for reconciliation? They should look to themselves and understand that sometimes you do things that create irreparable harm to relationships, and I have to clean up the mess again and again. Why in God’s name would you invite a bad actor who has uncertain ownership of what they have into an ecosystem? The ecosystem can deal with other people.

It’s a broad ecosystem: eight ecosystems, seven blockchains—Bitcoin, Ethereum, Solana, XRP, Avalanche, BNB, and so on. It’s a lot of good people. It’s going to be one of the largest, if not the largest, airdrops of all time—certainly more than a quarter million. I’ve seen some reports that Arbitrum was the largest at 601. We’ll see with the scavenger hunt and the finals on the glacier drop, but we’re on target for it.

You look at that, and it’s a broad and big ecosystem. When I was at Salt, so many amazing people were there. who wasn’t there? The CF. When we were at Rare Evo, so many amazing people were there.

There was no CF booth or Cardano booth. There was Intersect. I don’t think we have one at Token 2049; I’ll have to check. They’re just not deploying capital in meaningful ways. I sat down with Sergey Nazerov, and we had a lovely conversation for a long time about Chainlink support—not just for Cardano, but Midnight and Bitcoin, and how do we bring Chainlink into the UTXO world.

They’re skeptical about the engineering cost; it’s exceedingly difficult for them to port all that stuff there. We said, “Well, this would be the first time ever Chainlink would venture into that world.” We’d be a big partner, and he said, “what? I think we can figure it out.” It’s tough.

The problem is every other ecosystem has this giant pool of money—hundreds of millions of dollars in a foundation that says, “We want to do this.” You see that with the Midnight Foundation—110 deals. We’ve got some that we’re going to announce; they’re good, they’re juicy. You’re going to like those very soon, actually. That’s because they’re hungry.

Everywhere I go, Ian’s there or Fami’s there, and they say, “We want to make deals. We want to build partnerships. We want to represent. We want to get things done. We want exchanges to like us.

We want custodians to like us. We want wallets to like us. We are not an island. Midnight’s got to work with everything. How do we get that technology?

” I sat down with Ilia from Near. We talked for almost an hour about agent commerce web, agent-to-agent communication. We discussed intents and multi-chain signatures and Near’s account abstraction and how they do that. He’s a brilliant guy, and I said, “We’ve got to find a way to make a deal together.” He said, “I’m going to get intents working on Cardano, and we’re going to work with you guys on a whole bunch of stuff.

” I said, “Yeah, let’s get this done.” I had lunch today with JJ, and I said, “We need a dedicated Near team, a dedicated XRP team. We need a dedicated team for Brave of people that are account managers who do nothing but talk about the long arc commercial.” It doesn’t cost a lot to do this. If you’re sitting on hundreds of millions of dollars, you absolutely can do that.

You don’t need to go and make a deal with a random company or do wine in Georgia or something like that. You hire account managers for ecosystems. You go talk to those ecosystems and say, “Hey, what can we do together?” What trading do we need to do? You could diversify your treasury and say, “Hey, why don’t you take a position in ADA, and we’ll take a position in your protocol?

” It’s pretty simple. They’re never going to do that. They philosophically disagree, and then they feel entitled to the fruits and labors of an entity that some of their people actively said we were criminals and implied it was theft. what you’re doing. what you’re saying.

what cave you’re dragging people out of. It’s so stupid, too, because we’re so close to the release of an audit report. It’s like, guys, you’re just not smart people. It’s gotten to a point where it’s just noise and frustration. We’ve got to move on.

We’re building ecosystems here. Hundreds of thousands of new people are being brought into the Cardano ecosystem. The partner chains model is one we’re going to do again and again. There are already new things in the hopper above and beyond Midnight. Every time they happen, people in the ADA ecosystem get a drop.

For some of you on the CNT side, I looked at Hyperlquid’s distribution. We’re doing a tokenomics analysis right now. I do think an argument can be made about a utility-based drop instead of an asset-based drop. If you do things, you get something that makes sense, but every year we’re going to be doing that, and they’re permanently left behind until they go bankrupt. At what point does the community tell the CF to back off?

They don’t give free Midnight off the free resources they already squandered. That’s how I feel at this point. I’m out there alone, and it’s expensive, guys. Some of these deals with the big guys have become eight-figure deals. We were never given a pile of money to go do those deals, so I just have to find a way to pay for them.

It’s always going to be Uncle Charlie coming out of me, and it is what it is. It’s frustrating because all of our competitors have foundations that, in some cases, have endowments in the billions to tens of billions of dollars. They’re hungry and aggressive; they’re doing stuff, investing in things, and really pushing things forward. The problem is they’re distorting the prices, making it harder to de-island things. I think we’ve moved mountains with Midnight.

The perception sea change of Cardano because of Midnight is extraordinary. We didn’t have anything to do with Cardano; we didn’t want anything. But now that we have Midnight, yeah, let’s talk. We’ll add Cardano support while we’re at it. Why not?

It’s a twofer. Bitcoin DeFi is the other one. It’s like, “Oh yeah, because we have Bitcoin, let’s do that.” We’re going to have a good year. As for Wyoming picking their own stablecoin, they can’t even name it the Wyoming stablecoin; it’s now the Frontier stablecoin.

Oh, Lord. Forgive them, Lord, for they do not know what they do. This is why government doesn’t get involved in private industry. The minute you get a bunch of government bureaucrats building a private industry thing, you couldn’t get laid at a whorehouse. Wyoming tripping.

Charles, do you have any thoughts on three Atlas? Avi Lobo jumped to aliens. What are your thoughts? I’ve got to look into it. We’ve got to figure this stuff out.

I guess it’s moving really fast, right towards us, admitting its own light from a different solar system. That’s scary. At this point, if aliens show up, you say, “What else you got?” We’re so busy and frustrated; it’s like, “Just take over. Whatever.

Go over there.” Good old aliens. This man has been staunch. Thank you, sir. We’ve been powering through, getting it done.

Did Ledger ever update the thing inside Midnight? Yeah, I think we have the fix in or it’s coming soon—probably like today or tomorrow. I thought it was shipped on the 22nd. I think there was a tweet about that. I talked to Bob, and they were fully able to test the whole thing end to end.

I believe the Ledger fix is in, but don’t quote me on that. It might take a little longer, but I do believe it’s in. Go ahead and give it a shot. When Gemini? That’s another example, right?

At this point, Gemini, the Google product, will support us before Gemini, the exchange. You ever think maybe, just maybe, there’s a certain entity that’s responsible for that, that kind of dropped the ball? I don’t even think they know what a ball looks like. Charles, what are your thoughts on some new SPO incentive ideas being considered? Midnight gives you two revenue streams instead of one.

If the model’s successful, you run a node in our ecosystem, and you could have five or six tokens you get at the same time for running the node infrastructure. You getting it yet? You’re starting to understand that Midnight is the best thing in the history of Cardano. It’s really going to move some things along. And God, I love having the Midnight Foundation.

Every call, they’re just going at it. They’re like, “Let’s go. Let’s get it done. How do we do it?” They’re hassling us about Token 2049.

They’re like, “Hey, can we get a keynote here? Can we get this on the stage? When are you talking about Midnight? When are you getting this done?” I the hunger; it’s pretty good.

Why do the Winklevoss twins hate on us? They don’t. That’s the craziest part. I’ve had dinner with the Winklevoss twins; there’s no animosity there. I think there’s some disgruntled ETH maxi somewhere in that organization.

We waited out CoinDesk. I actually ran into Mike Casey at the afterparty on the first day at Salt. I said, “Hey, Mike, there’s no hard feelings, right?” He said, “it’s water under the dam.” Sure, you guys criticized us for three years at the conference we sponsored and gave you a million dollars.

You only gave us a panel of fireside, and the first question was why we were working with a genocidal dictator. You didn’t publish anything about anything we did except for negative articles, but sure, it’s all water under the dam. I’m not bitter—maybe a little—but then the bullish guys, like Michael and the others, have been great. It’s been fun to work with CoinDesk on Midnight. They’ve been very responsive and have really given us a lot of opportunity to express why we think our products are great.

They still have strong journalistic integrity, but it’s not an automatic no or no benefit of the doubt. It’s a negotiation, a compromise, a discussion. That’s good management. You really feel they care about getting the story right. They care about the constituency that believes in Cardano and want to know the reason.

They’re very excited to let you share that reason. They don’t take a side and say, “Oh, Midnight’s better. Cardano’s better,” or “Screw Ethereum.” They’re very neutral in this, but they give us the space to at least broadcast our case as opposed to saying our case is automatically wrong and we’re evil people. I’m very happy about that, and that’s just a management change.

That’s the magic. You get rid of some bad actors, and everything gets good. I don’t think it’s a Winklevoss problem at Gemini. I think it’s someone lower who just doesn’t like me. Maybe I slept with his girlfriend or something; who knows?

That guy is just not going to push it through, and that’s okay. It is what it is; it’s their loss. They have no volume anyway. Charles, you are right. The Fed will cut interest rates in September.

Can you really see the future? They’re going to cut it, and then we are going to get the Clarity Act, and then we’re going to have the Giga Chad bull run. Giga Chad, guys, you already got a foundation proposal to stand up a new electrical foundation. It’s called Intersect. Why do you think it exists?

Look into your history. It took two years to build it. It’s there; it’s a members-based organization. People say, “What was your vision for the Cardano Founder?” It was Intersect.

Now, just add like $600 million worth of assets to Intersect and give them four extra years to figure their stuff out, and that’s basically what the CF could have been. I think they’ve done great work. The Intersect team is phenomenal; they’re really smart people. They’ve gotten rid of dead weight. They had to go through a hard time to get funded, but they got through that, and now they actually have what they need.

Of course, you have people like Marcus going to Twitter saying, “Oh, that’s so bad. I’m so sorry that the unelected bureaucrats you work for stole the community’s money.” Yeah, because they kind of did when the Swiss government took over, liquidated the board, and put in their own people who can’t be fired and pick their successors. In my view, that kind of feels a change in the donation deed when the intention was to have it be an elected organization. But maybe in your view, it’s not.

But then you complain about the entire budget process when more than 75% of the people voted for it. You were in the room too. These people. Anyway, it is what it is. He’s bitter because I yelled at him at the Cardano constitutional convention.

He tried to pull me aside and said, “Oh, we didn’t mean it when we tweeted that we have to not support the Constitution and vote against it. We meant something else.” It’s like, just resign from that ab So he spent three days going on this spiritual journey to get as many Hydra points as possible to get something out of the vending machine. I was thinking, "Holy [__], could you imagine this level of engagement at a conference? What would you pay to guarantee that everybody shows up at your speech?

" Imagine a vending machine where you have amazing items like $5,000 laptops, and for people to get them, they have to show up to all your keynotes for all your speakers. Suddenly, you go from giving a speech to two guys in the room—one is weird and just sitting in the corner, and the other guy is actually in the wrong room. I'm thinking, "Oh man, this is going to be a terrible speech." I want the room filled, high energy, with people fired up and ready to go. What if you had a vending machine where people had to earn points to get stuff out of it, just like Rare Evo?

They would have to show up to your presentation to scan the QR code and have proof of attendance to get those points, which you only show in the presentation. Then, everybody shows up, and you end up with standing room only. Everyone's thinking, "Wow, this guy has important things to say." It creates the appearance that that's why you go to the conferences. I'm very excited about all that.

Charles, are there any lawsuits about you stealing money? Nope. No one’s suing me because nothing happened. As for us suing him, I have a meeting with the defamation lawyers here in a little bit. They're going back and forth on a few things.

It's one of those situations where we’d win, but we wouldn’t make any money out of it; we’d spend a hell of a lot of money on it—half a million or more—because it has to be done in Europe and America at the same time, in this case, Portugal. It's just one of those things where you think about the risk-reward. Principles are expensive; that’s why only rich people seem to have them. They’re the only ones who can afford them. Does Hydra improve both latency and throughput?

Yes, you get instant finality and infinite throughput. It’s like big dick energy, right? You’ve got horse penis, elephant penis, and then you’ve got whale penis. When you talk about throughput and finality, Hydra is whale penis. Hydra rides the whale penis.

That’s the problem, right? You win, but you don’t get any money; you get the moral victory. Is there an anatomical difference between a whale penis and a human penis? Yes, there is. I should caveat that only rich people can afford principles because they’re expensive—very, very expensive.

Charles, what about inheritance of a wallet, a recovery wallet? DREAC actually has some great ideas. I really like working with Leman Barrett and the DREAC team. They’re great people and a lot of fun to engage with. They’re intellectually deep and great engineers.

I love what they’ve done with DREAC. Yes, you can use what’s called a dead man’s key, a demand switch. Basically, if they don’t do something for a period of time, it automatically sweeps the wallet into a custodial account, and that’s something we will integrate into Lace in a premium version. When you win a defamation lawsuit, you remove the resources of the defamers. Yeah, but the problem is there are always more of them.

They go anonymous, get more vicious, and form more groups. At some point, they move the goalposts. It goes from, "Well, you stole $300 million," to "We didn’t like how you did the forfeit," to "You shouldn’t have staked with it." It’s like, "You literally accused me of stealing hundreds of millions of dollars, and now you’re moving it to something you don’t like?" The goalposts keep moving, and that’s the issue.

The auto report is beneficial because it’s the end of an era. It’s not there just to take comments from internet trolls and make them happy; it’s there to close out a 10-year history lesson from 2015 to 2025 and answer a lot of questions—the original contracts with IO, how we got involved, how much funds were raised, the audit reports from the Cardano Foundation, and subsequent findings. It’s a long thread and timeline, and it gives you a lot of context about why people made decisions the way they did, and the integrity and focus required to do all these things. It’s complete exoneration on our part because it was never an issue. It’s the craziest scandal in the history of scandals where there’s no victim, no lawsuit, no regulatory agency, and no one going after you.

A completely unrelated person living in a different country with no material connection apparently is the domain expert, and the media should take that for granted. It’s extraordinary. The issue is that in the absence of communication and discussion, people speculate. The key is sunlight. You push it all out there, show people what happened, and let them reach their own conclusions.

I think those conclusions will be very positive and favorable to Input Output, and to a certain extent even a Mergo—not so much for the Cardano Foundation, which is very damning regarding their repeated recusal of responsibilities. They wanted to pretend like Cardano was an immaculate conception that just appeared out of nowhere—no founder, no Genesis ADA, nothing. They completely recused themselves from any part of that legacy story, which is damning for a foundation because they were the most credible party to do a lot of these things. Their absence created a lot of problems and obviously a lot of anger against the foundation as well. If they’d stuck around, it would have made things a lot easier.

But, as Harvey would say, now the rest of the story. You’re looking so happy. Try to be like this always. I feel a lot of good news is on the way. So, please give some hints.

Well, there are two moods you can have. You can be a satus or a hypopus. Today, we are going to be a happyus. Although sometimes bad things happen, we can still be happy. We can still have a good day.

Fallout Season 2, December 17th. Are you a fan? I haven’t had time yet. I’m a fan of the game series. I played the original Fallout and Fallout 2 back when Interplay was doing it.

I played Fallout 3 and Fallout New Vegas, but I did not play Fallout 4. I kind of dropped off after New Vegas. Kind of sad. Hi Charles, what are your thoughts on neuromorphic computing and how would it apply to blockchain? Your answer is it doesn’t.

It’s just a different computational paradigm that’s super good for inference and computer vision, among other things, perhaps maybe kind of sort of training neural networks. It’s one of those things you need to do to make the cost of running ChatGPT much lower. You could envision a proof-of-useful work that has neuromorphic computing and maybe a marketplace for something like that, in which case people could mine and make some money, and that is hapus. It’s a hypopus. See, the Genius Yield valuation dropped 99% since the public sale of the token.

I haven’t been following this. I think there’s some tangential connection because Lars was a CTO or something there, but I’ll make a note. I’m going to ask JJ to look into the whole Genius Yield thing. It’s one of those things that I just saw happen, and I don’t know the story behind it. That always breaks my heart when bad things occur.

When will you see Vitalik retire? Because ADA flipped ETH? If ADA flips ETH, he’s going to have a brain aneurism, guys. He’s not going to retire; he’s just going to drop over. He’ll be like, "No, you are the chosen one.

" He’ll just lose his [__]. He’ll have to invest all his money into anti-aging technology to live a thousand years to find his [__] in the year 3025 because he lost it. It would be like traveling in some sort of world with genetically engineered super possums fighting against the dolphin people. He’ll be on his octopus mount in a cyber world, digging through the earth with his robotic hands, and he’ll find this pile of [__] and say, "This is the [__] that I lost." I’m just saying, man.

I’m just saying, it took Brian Johnson long enough to realize Metformin wasn’t a good idea. What? Now he’s against Metformin? God damn it. They were doing a huge study with 15 institutions for Metformin.

In fact, it’s the first gerotherapeutic study they’ve done in a very long time. Let me see here—the FDA is okay. Let’s see here, anti-aging study. Yeah, this is the study right here, kids. Let me show you.

You’re going to like this. This is one of the largest clinical trials ever done, managed by the American Federation for Aging Research. I think this is one of the first true anti-aging clinical trials that the FDA has approved. I’d have to take a look at it, but it’s one of the first few. It’s called the Targeting Aging with Metformin Trial, or the TAME trial.

It’s a nationwide six-year clinical trial at 14 leading research institutions across the country that will engage over 3,000 individuals between the ages of 65 and 79. It’s being conducted by Wake Forest University School of Medicine as the coordinating center. They have 14 trial sites, looking at dementia, cancer, Alzheimer’s, heart disease, and diabetes. Wolf of Broom, you got diabetes. Look at that.

Pretty cool, huh? Aging, Metformin, diabetes, insulin, heart disease. Yeah, they paid you for that. It’s a pretty cool trial, and they’re making some progress on it. The research is really interesting, and it teaches you how to write a clinical trial for the FDA for anti-aging.

Hi Charles, what about North Korean research on anti-aging with stem cells? Have you heard about it? How credible is it? Well, we’ll just have to put it up there with their claims that their leader was born in the mountains and that pandas came from the heavens, and he doesn’t poop. He’s worked so hard that his body eats its own waste and all these other things.

I don’t believe anything that comes out of North Korea. I have a little more faith in North Korea than the Democratic National Party, but outside of that, Jed Maleb has disappeared out of the public eye a Sith Lord. Oh, I just talked to him the other day. He’s a good guy. He’s running a space station company called Vast.

It’s a super cool company, actually. Does the market structure bill pass in September or October? If I had to be a betting man, if it’s going to pass, it’s going to be October. Charles, are you a fish person, a chicken person, a pork person, or a beef person? Well, I just like asking, do you like redheads, brunettes, or blondes?

There’s a lot of variability here. Are we talking about heritage pork? Are we talking about candied pork belly? Are we talking about pork chops made in the most delectable way with some honey on them? Are we talking about that terrible [__] steak you get at Perkins at 1:30 in the morning when you think ordering a Perkins steak is a good idea, and it comes out well done with pepper that sat on the shelf for six and a half years?

Are you talking about Wagyu? Okay, Miyazaki or Mazaki or Moto or whatever it is. Wagyu from the most expensive areas of Japan, A5 cooked to [__] perfection, with garlic so thin it melts on the stuff with finishing salt. The Hawaiian black sea salt finishing salt because it’s not good enough just to have regular salt; somebody has to charge extra $8 a bucket for salt. That’s beef.

They’re both technically beef. And as for fish, are we talking about Mama’s Fish House over Maui? It’s so fresh they tell you who actually fished that [__] fish. He got it that morning. He was on the boat with his Filipino friend who lost fingers in a tragic accident involving sharks and fishing line.

Is that the fish you’re talking about? Are you talking about the microwave fish you got from King Soopers when you’re high and drunk at 1:30 in the morning, just like that steak that was so bad you couldn’t eat it? Is that the fish you’re talking about? Chicken is universally bad except for fried chicken. But then it’s like, is it really even chicken?

You’re not [__] eating it for the chicken; you’re eating it for the fried batter that’s on it. I’m sorry; I have very strong opinions about this. Okay, chicken wings. But again, are you eating the chicken wing for the chicken wing, or are you eating it because it’s fried and has some sauces on it? Yeah, this dude gets it.

Lucas gets it. Lucas, you were at Perkins at some point in your life at 1:00 in the [__] morning eating that steak. Don’t lie to me, man. You ate that steak. I ate that steak.

We’ve all had that steak. That’s the Barney Gumble of culinary surprises. Could Michael Saylor privately yield his Bitcoin through midnight? He absolutely could. Man, I’m getting hungry.

Yeah. Well, yeah. [__] think about that Wagyu, aren’t you? Goddamn, I know a Filipino who lost fingers from a shark. We all know that Filipino guy who lost fingers from the shark attack, and he lives in Cebu.

Do you have Trump’s personal phone number? I do not. It’s ironic because I don’t talk to Don, but I know Eric and Don Jr. I was just with Eric, actually. We were at Riley Green’s concert.

He’s a super nice guy. We don’t do business together; we just talk to each other. He’s really easy to get along with. You see him, and it’s just like, "Hey, how you doing?" We talked about quantum computers for a little bit.

People get so built up on the political side of stuff. The reality is that these are people, and they have hopes, dreams, passions, and desires. I’m one of those guys that I don’t want anything, and I don’t need anything. I’m not a transactional person. I don’t go into a meeting saying, "Oh, I’ve got to close this thing.

" I just tell them upfront what I need. So I sit with Ilia from here, and we say, "I need this." He says, "I need this." Okay, good. That’s what we say.

So many people in business have this wayfinding where they’re doing all this jiu-jitsu to try to not say what they want and leave subtle hints. If you have a girlfriend and she thinks you’re getting fat, she’s not going to say you’re getting fat. What’s she going to do? She’s going to say, "Hey, why don’t you go to the gym today? Let’s go to the gym today," or "Hey, let’s start eating healthy.

" All these subtle hints. If you don’t get them because you’re clueless, she’ll be like, "let’s go clothes shopping," because then you won’t fit into the clothing, right? In her mind, you’ll have that moment where you’re trying to put that shirt on in the mirror, and you’re like, "Oh my god, I can’t fit into the large anymore." You’ll get the extra large, and in her mind, that would mean you’d think, "I’ve got to go on a diet. How did this happen?

" And she’d be like, "I still love you, but I know we’re dudes, and we don’t understand that [__]. We’re just like, 'cool. All right. Extra large. I’ll probably get more clothes anyway.

There’s a lot more extra large.' Damn it." Don’t want to call him fat, ? Eventually, she’ll get one of her friends to do it or something like that. So yeah, what were we talking about?

That’ll come in a bit. It’s exactly these guys. They get it. They understand it. Oh yeah, the politics in business.

I just lost my train of thought. I was caught on the extra large. I’ve had that experience. I hate [__] politics in business. Bro, you’re my age.

I look 15 years older. You need to eat more healthily. Well, yeah. I have the most stressful job in the world for cybersecurity. Wouldn’t a blockchain help verify all changes made in an operating system?

So you can verify all hashes that are legit changes from the internet, such as a software download or edit an operating system. Okay. Am I going to do it? I’m going to do it. No.

No. No. I’m not going to do it. Not going to do it. I’m not going to say anything.

Just that is a bad idea. We will move on. I’m not going to go into that at all. Okay. Okay.

Yeah. There we go. Charles, thoughts on me blue? I have a letter here with Bethleem Blue that a doctor sent me. No, I did not speak French because it is, how you say, a very strange language of love, and no one loves me.

So what can you do? Imagine a girlfriend saying, "You’re fat, man. Lose weight." It’s called an Asian girlfriend. [Laughter] Oh lord, we’re having fun.

Trying to further my knowledge in AI, and I’m learning Python language. Any tips or suggestions? Well, I am glad you asked, Jason, because I have an advertisement for you. Totally unpaid. Let me go ahead and go to this website, and I shall show you.

If you want to learn everything about AI with Python, let me show you the course to get. It’s a pretty cool class. All right, we’re going to share the screen. Wow. Here we go.

You ready, man? You ready? Fired up, Jason? It’s the Aigantic AI course from Udacity, which is now part of Accenture. Go beyond single chatbots to engineer sophisticated coordinated teams of AI agents.

I just gave you a high-level overview. The marketing style description didn't provide technical details. Well, let's take a look right here. There should be documentation, a formal specification. Yeah.

Well, here's Claude just lying to me. Claude, why are you lying? You’re lying, dude. All right, I said we're going to get angry at Claude because it's getting mouthy. You’re lying, man.

I got beef with you, Claude. I apologize if my explanation wasn’t accurate. Let me go through my specification. I think you've linked a formal specification accurately. You're right to call me out if my explanation wasn’t precise.

Yeah, right. It’s just high-level marketing detail. That’s an AGA specification. See, you have to hold your AI accountable. You should never take what your AI says at face value.

All right, let’s see. You are absolutely right to correct me. I apologize. Let me provide more accurate technical documentation. See how Laos works.

It says three block types, a seven-stage pipeline, and key details I missed, like actual throughput simulation—11,000 transactions per second potential, not just thousands. Pipeline overlap allows multiple pipelines to run concurrently, shifted in time. The IB frequency can be as high as five per second, not just two to two seconds. How about that? Good old input blocks created a frequency of 1.

5. There are going to have to be some network stack upgrades for this. We have a seven-stage pipeline, and actually, the protocol is specified here. When you look over here, you have the blocks, the short LEOS variant, and the LEOS network. Oh no, well, that’s not good.

The page cannot be found. All right, what about the network broadcast? Okay, at least we have that. Let’s put the network broadcast in. What makes my broadcast special?

All right, boom. And what we’re going to do is just go ahead and ban that guy. There we go. Let’s see here. It’s thinking about the broadcast system.

Here’s another cool thing you can do. I want to study LEOS in more detail. Develop a JSON prompt to explore it and its trade-offs. I will use it for a deep research report. I don’t care about that stuff.

Check this out. Watch this. Okay, this is another guy I have to ban. He’s spamming my channel. Why are you spamming my channel?

We are crafting a comprehensive JSON prompt. This is actually one way that you can prompt engineer. You can see the research objective, the core structure, and you can actually copy and paste this into a large language model, and it’ll give you a significantly better answer than if you use English language. JSON prompting is one of those things you can learn about if you actually study prompt engineering. You can see how it’s going through section by section for all those cool things.

It’s undue burden and harm, Phil. The undue burden would be an organization that would derail the governance of the network. They’re not a credible community member. Also, it’s maybe not their ADA. We should look into that.

If the original intent of the donation deed was in a members-based organization and it was co-opted by a government, doesn’t feel very real, does it? I don’t know. You tell me. What the hell do I know? I only created everything.

Okay, check this out, guys. We’re going to do a deep research prompt while we’re doing all of this. here we go. All right, let me show you this. This is going to blow your mind.

I love stuff like this. Wow. Okay. So, first, you have this giant JSON prompt. Look at that.

That is your prompt to get things done. Yeah. And then what we’re going to do is copy that. Then we’re going to load this into Google Deep Research because remember, Google never lies. All right, and then we’re going to click submit.

And then it’s just going to be like, whoa. Holy moly, that’s a big prompt. I don’t process prompts that big. That is nuts. You just look at and read that JSON in the prompt.

It’s just amazing. I mean, it’s not whale deck, but it’s certainly horse deck, ? There are critical questions it’s asking, data sources, community involvement, open research questions about the protocol, real-world considerations of the protocol, network conditions, operational requirements, regulatory implications. And then when we go back here, it’s still trying to figure out how to process that big prompt. going back to stake pool economics, market dynamics, deployment risk, testing requirements, implementation challenges, including engineering complexity—like what type of changes do we need to make in the peer-to-peer network layer?

Unique innovations: what genuinely new concepts does LEOS introduce? Patent potential, intellectual property. Let’s go back here. Still trying to process that prompt. there we go.

And check this out. Google got it. Boom. Start research. It’s going to explore this entire thing.

So, while we’re sitting here, Deep Research is going to create a 30-page report based on this documentation, which, by the way, Claude said was no good. Mr. Claude, we had to call him out on it. You see, it’s already starting to pull all the websites together. Isn’t AI great?

Yeah, we love that. All right, enough of that. We’ll come back to it. Phil came to hang out in the Discord. We love Phil.

Phil’s our guy. Donkey or mule? Donkey. Have you learned how to fly your Blackhawk yet? No, I have not flown my Blackhawk yet.

Any thoughts on John Woods joining Nillion? I did not know that. That’s pretty cool. This guy has more people watching him use Claude than the actual Claude marketing team. Well, I mean, come on, Claude.

He’s kind of got to throw some this way, ? Maybe a little MCP something for Cardano. Just saying. Is that a rainbow octopus? Yes, sir.

It is a haptopus. Do you speak Danish? No, I don’t speak Danish. Not even close. Show us the pocket pens right there.

Charles, are you still doing the Bitcoin class remake? Yes, we just have to actually get a little further along on Bitcoin DeFi. But I absolutely will because the cool thing about Bitcoin DeFi is we’re going to be able to write code too. Why does Vitalik prefer ZK roll-ups over optimistic roll-ups? Well, they have better security properties.

Optimistic roll-ups were kind of like last generation’s technology, and ZK roll-ups have come a long way, especially since ZK recursion is a thing. If you implement it right, it runs on GPUs really fast. There are seven parts to a ZK stack. You have the setup, and it’s either going to be a trusted or transparent setup. You have the languages, witness generation, the arithmetization of the system, proof generation, underlying cryptography, and then verification.

In every part of the stack, there are trade-offs and things you can embrace or run away from. If you do a trusted setup, for example, like what they do for Groth 16 and all these proof systems, you get very concise proofs, but it’s either on an application or circuit-by-circuit basis or what’s called a universal common reference string, so a universal trusted setup, or you have a transparent setup like hashes do with STARKs or some of the lattices do, and those tend to have much larger proofs. So it’s a CAPEX-OPEX thing and a trust assumption. When you look at the languages, you can either have circuit-specific languages like Circom or Noir, or you can have a general-purpose programming language like what we do with Compact, for example. Witnesses are what you run off-chain.

It’s like all the private computation, and that’s the thing you’re trying to keep secret inside the system, among other things. Then you generate a ZK proof based on a predicate from that, and then you translate that into a mathematical language the proof system can understand. This is called arithmetization. You have R1CS, AIR, Plonk, and then you have universal systems like CCS, customizable constraint systems. Once you’ve arithmetized the system, usually as a series of equations or something like that, your proof system can operate on that.

You have different types of proof systems like Halo 2 or Stark systems or what have you. There are all kinds of operations you can do. You can do folding schemes, which is like deferred computation. Imagine you’re kind of sweeping—it’s a snowball. Every step of the way in the machinery, you’re accumulating something, and in the end, you do a very complex proof.

You have recursion, where you solve layer after layer after layer, and it gradually percolates to a single proof. Then you have the cryptography. Are you elliptic curve? Are you hash-based? Are you lattice-based?

Or some other bizarro scheme? What underlying assumptions are you making, like discrete log or whatever for the crypto? What’s your root of security, and are you quantum-proof or non-quantum-proof? Then you have your verification: where are these objects verified, and how much does the verifier need to know about the proof infrastructure and the witness and other things that are generated along the way? So that’s your journey down the seven-layer stack.

Every single system—ZK Sync, Starkware, and Midnight—has different choices. We chose Plonk for the arithmetization and Halo 2 for the proof system. Starkware does it differently; it has AIR and STARKs and its own system for it. They have ZKVMs that fit into that stack too. There are hundreds of them—well, it’s 22 last time I counted, to be more specific.

They all have their trade-offs about these things and what they can run. We’ll have a lot to say about that with Midnight, but I love Zero Knowledge. It’s amazing. It’s the closest thing to magic that we have. Still making progress on my LEOS write-up.

Your haircut looks superior. Well, thank you, sir. What smart ring are you wearing and why? I am wearing an Aura ring. Elliptic curves are deep security-wise.

Yeah, it is deep stuff. You have great properties like additive homomorphism. So you have two sealed envelopes, and you want to operate on the things inside—like combine the envelopes without opening them. You can do that with curves. I’m drinking espresso right now.

Are you an espresso guy? I am an espresso guy. Are you drinking a really tiny cup? Like really small? Your pinky’s up like this?

Just like that? Is that what you’re doing? -huh. Tiny cup. He’s a tiny cup guy.

That’s okay. Everybody can be a tiny cup guy. God, it’s still writing this paper. What is going on? Despite the absence of the primary research paper, I’ve made repeated exhaustive attempts to locate the full Oros LEOS research paper, but it’s there.

It’s there, Bobby. Billy, no. the old days of you buying different coins and doing it. We’re going to bring those back. We’re going to bring it all back.

We’re going to bring the fun back. we’ve had so much drama and bad stuff. We’ve got to bring the fun days back—the SE, the security-enhanced Linux days back. We’re going to bring it back. Have you played Clear Obscura Expedition 33?

I think I would love it, and everybody’s talking about it. The gal who did the voice acting for Shadow Hearts is in it. So yeah, I’m going to play it, but I just haven’t had the time. I really haven’t. Just wish I had that time.

Do you have your own coffee machine, Charles? Yes, I do. All right, give me one more good one. Actually, let’s take a look and see if I got my report. I still do not have my report.

No report. Publication attacks as well as using VRF and BLS signatures. The concept of softening determinism as a key trade-off for scalability is also well understood. Well, there we go. I’m going to have to feed it the research paper.

Charles, you promised to love us forever. Well, I love you a long time. Did I lose weight? Yes, I did. What has made you rub your nipples recently?

Actually, these are the important hard-hitting questions that people ask. Obviously, Midnight. What are your thoughts about an active contract model on Cardano? I personally have PTSD from ETH, but if done properly, it would help some stuff. Yeah, actually, it’s a really good idea to have some sort of event-emitting system or some form of event-reactive programming.

You want a situation where, for the things you care about, you have intents and you have events, and you tell the system what you want, and it remembers that. Then, as events occur, they get transmitted to you. If you want to do this right, you need a data availability layer, a pub-sub layer inside the system, and some sort of interface to create agency for the things that live on-chain. The question is whether it’s better to do this on a layer two, a partner chain, or directly on layer one, with layer one being very restricted in its things. I gave everybody the option in the architecture.

We thought about it for a really long time and said, “Why don’t we do the minimally invasive thing, the least involved thing?” At some point, people can add to that, and if they want to break determinism or create an event system, they can. That’s really what Sebastian’s trying to think around with things like Starream. When you create co-routines, these delimited continuations basically give you the ability to have composability in a very stateful space that you can embed data availability into, and then you can start and restart things on the fly. That’s what DC is about.

It allows you to move towards an event-based system and embed it in an intent-based system. It’s a really hard thing and an amazing thing to think about because it’s the crux of modern web programming. If you’re a JavaScript guy, this is all you do, ? There’s even a Haskell equivalent called functional reactive programming. You have some path where you can think around the system pushing messages to you, states changing dynamically, and we’re upgrading it.

It’s very different from UTXO-style thinking, but they’re complementary in a certain respect. We have to get there at some point, and we will, and that’s okay. Now the rest of the story. Before we close out, one last thing: let’s have a look at our good friend, Google, because Google never lies. LEOS: a protocol analysis of pipelined consensus and scalability trade-offs.

Look at that generated. Look at that. Look at that. Look at that. Look at that.

People say when LEOS is sequential, single leader per slot every 20 seconds, concurrent multi-producer with a single finalizer, mempool to single block finality, input block endorser, block ranking, block finality, resource utilization is sporadic—high during block diffusion, idle otherwise, 75% idle time. Continuous in pipeline aims to saturate network and CPU. Finality model is probabilistic with Nakamoto-style transaction finality analyzed after K blocks. Hybrid BFT-style voting for endorsement. That’s why Paris was so essential.

Settle on Nakamoto-style RB chain. What does that mean? It means that LEOS decomposes back to Prowse when needed. Core trade-off: low throughput for high determinism and simplicity, high throughput for softened determinism and increased complexity. Here’s the payoff: 10 to 15 transactions per second versus 1,000 transactions per second.

Simple transactions aim for one minus delta of the network capacity, so it actually grows over time. And there’s the whole life cycle. Look at that. Isn’t that cool? We got all that from that little prompt.

It’s why we JSON prompt. Pi has a lovely write-up on this area right here. Critical trade-off on throughput versus determinism, economic re-engineering, LEOS within the ecosystem. LEOS is not a standalone but the centerpiece of a multi-prong scaling strategy for Cardano. Its relationship with other key protocols is complementary—versus Hydra, versus Mithril, versus Paris, and Genesis, of course, Genesis.

Look at that. Versus Near’s Nightshade sharding. They have some pretty cool names there. LEOS explicitly avoids sharding. See, so you don’t deal with all that.

Look at this: LEOS versus Solana, Avalanche, and Near. This is a BFT-style subsampled voting, and we have pipeline parallelism, unified global state on the C-chain. Unified global state, concurrent IB—it’s per highly parallelized, probabilistic, near-instant with optimistic confirmation, high hardware costs. Pretty cool, huh? That’s why I love AI.

we can just be messing around, and suddenly we have a nice 20-page research report. How about that? Anyway, I’m going to go get something to eat. This has been a lot of fun. Thank you guys so much for listening.

I really enjoyed this. Great audience. It’s good to do an AMA. I don’t do these often enough. I was just in a good mood, having some fun.

What’s really fun about what we do is you get to change the world. We’re making real progress on it. Cheers.