Steve Dircks

Guarantee he will be the scapegoat for WLFI in the coming years...

@justinsuntron @pangramlabs AI?

This Is World Tyranny, Not World Liberty Financial — Here's Why This proposal has been packaged as a "governance alignment signal" and a "long-term commitment," but strip away the packaging and what you have is one of the most absurd governance scams I have ever seen. Let me break it down. I. Vote Against and Get Punished — Classic Coercion Tactic This is not a legitimate voting exercise, not even close.  The design of this proposal is a logical trap: anyone who votes against it has their tokens locked indefinitely with no unlock path whatsoever. In other words, if you oppose this proposal, you get punished. This is not voting. This is coercion. What kind of democratic process rewards agreement and imprisons dissent? II. Voters Have Been Selectively Frozen Out I personally hold approximately 4% of the voting power, yet my tokens have been frozen and I am forced out of this voting process. I am not alone. A large number of holders with significant voting rights are in the same position. Meanwhile, the team controls the power to freeze tokens — they decide who can vote and who cannot. What does this mean? It means the outcome was determined before the vote even began. This is not a governance vote. This is a performance where the police have already barricaded the doors of parliament and only let their own people inside to raise their hands. The voter pool has been purged. Only yes votes remain. The result of such a vote carries no binding force whatsoever. III. All Actual Power Has Been Seized by Anonymous Actors The actual control over the WLFI smart contracts lies in the hands of a 3/5 anonymous multisig and a single anonymous guardian EOA has the power to blacklist addresses holding WLFI. Let me emphasize — anonymous. This anonymous multisig can override any vote result and execute any operation directly at the contract level. The so-called governance proposals, on-chain votes, and community discussions are nothing but theater. Real power has never been yielded to anyone but themselves and it’s laughable that they try these tricks to fool the community. The bottom line is this, the power sits with anonymous wallet addresses whose owners nobody knows or can verify. This is not decentralized governance. This is dictatorship wearing the mask of a DAO. IV. Voters Must Identify Themselves, but the Rulers Are Anonymous — Worse Than Tyranny Here is the most ironic part: WLFI requires every participating voter to complete identity verification, electronically sign acknowledgements, and meet compliance eligibility requirements. You want to exercise your rights? Show your face first. But who are the guardian and multisig signers who hold the power of life and death over the contract? Nobody was told and there is absolutely no transparency.  While the governed must identify themselves, the governors with absolute power are anonymous. Your voters must register, submit to scrutiny, and be vetted and dictated how to vote— while your dictators won't even show their faces. V. A Naked Violation of Property Rights Worth Billions of Dollars Let us not forget the real stakes of this proposal: this is not some trivial parameter adjustment or protocol upgrade. This vote seeks to decide the unlock schedule for billions of dollars in assets, the reallocation of governance and vesting rights, and most extreme of all, the permanent destruction of billions of tokens. This is a naked expropriation of holders' property rights. In an environment where voting against the proposal is punished, where large numbers of holders have been frozen out of voting, and where actual control rests with anonymous wallets, using this sham vote to decide the fate of billions of dollars in assets? This is not governance. This is a sham and flies in the face of what this protocol was meant to be. No society governed by the rule of law would permit this. In traditional financial markets, any asset disposition of this scale would require rigorous regulatory review, independent board approval, and minority shareholder protections. Here, a few anonymous wallets get to decide everything. The permanent burning of tokens means holders' property is irreversibly destroyed, no compensation, no recourse, no due process. This has gone far beyond the realm of so-called "decentralized governance." This is a systematic violation of property rights. Conclusion I repeat that this proposal is not governance. It is an exercise of power by the selected few who are carefully engineering a further power consolidation and property expropriation operation. Dissenters are punished for voting no. Opponents are frozen out. Actual control lies with anonymous wallets. Those who exercise their rights must prove their identity while the fate of billions of dollars in assets is decided by a sham vote. This is not what decentralized finance was meant to be, results produced under these sorts of conditions carry no legitimacy, should not have binding force, and should not be recognized. I call on all WLFI holders to see this proposal for what it truly is, to voice their opposition across all public channels, and to reserve all legal rights of recourse.

@RobEducated @RealJimChanos cause we all started getting rate limited....

was early in seeing the "token" story but notice we peaked and past 2 weeks flattening out. @RealJimChanos $nvda $amzn $crwv $msft

@BoringBiz_ what about learning how to use AI?

Look, I totally get that AI is an incredible technology and maybe the most transformative one in our lifetime But it’s becoming clear that people are now trying to use AI as their magic tool to solve every problem, even when it has nothing to do with artificial intelligence > Haven’t read a book in 2 years? Try AI to summarize one > Don’t know how to code? Use AI to vibe code and build apps > Too lazy to write a thoughtful article? Use AI to write one for you > Can’t build a financial model? Have you tried using AI? The human brain is not meant to just immediately latch on to AI as the solution for every knowledge and skill gap that you have Learning things takes a lot of time and effort. Mastery is earned over the course of decades, and often involves starting with the most basic fundamentals It is amazing that AI can make some people 10x more productive, but the reality is that majority of the population lack any agency My guess is that this leads to massive cognitive declines in the future, especially amongst children who are learning to become reliant on AI I am not really sure what the solution is, but people should understand that AI is only one of the tools, amongst many, within your arsenal Put in the effort to learn new things and do not just rely on AI to do the work for you. It is a dangerous path to go down

@worldlibertyfi @justinsuntron @pangramlabs

Does anyone still believe @justinsuntron ? Justin’s favorite move is playing the victim while making baseless allegations to cover up his own misconduct. Same playbook, different target. WLFI isn't the first. We have the contracts. We have the evidence. We have the truth. See you in court pal.

@_LuoFuli @pangramlabs

Two days ago, Anthropic cut off third-party harnesses from using Claude subscriptions — not surprising. Three days ago, MiMo launched its Token Plan — a design I spent real time on, and what I believe is a serious attempt at getting compute allocation and agent harness development right. Putting these two things together, some thoughts: 1. Claude Code's subscription is a beautifully designed system for balanced compute allocation. My guess — it doesn't make money, possibly bleeds it, unless their API margins are 10-20x, which I doubt. I can't rigorously calculate the losses from third-party harnesses plugging in, but I've looked at OpenClaw's context management up close — it's bad. Within a single user query, it fires off rounds of low-value tool calls as separate API requests, each carrying a long context window (often >100K tokens) — wasteful even with cache hits, and in extreme cases driving up cache miss rates for other queries. The actual request count per query ends up several times higher than Claude Code's own framework. Translated to API pricing, the real cost is probably tens of times the subscription price. That's not a gap — that's a crater. 2. Third-party harnesses like OpenClaw/OpenCode can still call Claude via API — they just can't ride on subscriptions anymore. Short term, these agent users will feel the pain, costs jumping easily tens of times. But that pressure is exactly what pushes these harnesses to improve context management, maximize prompt cache hit rates to reuse processed context, cut wasteful token burn. Pain eventually converts to engineering discipline. 3. I'd urge LLM companies not to blindly race to the bottom on pricing before figuring out how to price a coding plan without hemorrhaging money. Selling tokens dirt cheap while leaving the door wide open to third-party harnesses looks nice to users, but it's a trap — the same trap Anthropic just walked out of. The deeper problem: if users burn their attention on low-quality agent harnesses, highly unstable and slow inference services, and models downgraded to cut costs, only to find they still can't get anything done — that's not a healthy cycle for user experience or retention. 4. On MiMo Token Plan — it supports third-party harnesses, billed by token quota, same logic as Claude's newly launched extra usage packages. Because what we're going for is long-term stable delivery of high-quality models and services — not getting you to impulse-pay and then abandon ship. The bigger picture: global compute capacity can't keep up with the token demand agents are creating. The real way forward isn't cheaper tokens — it's co-evolution. "More token-efficient agent harnesses" × "more powerful and efficient models." Anthropic's move, whether they intended it or not, is pushing the entire ecosystem — open source and closed source alike — in that direction. That's probably a good thing. The Agent era doesn't belong to whoever burns the most compute. It belongs to whoever uses it wisely.

@APompliano @pangramlabs whats teh verdict

OpenAI acquired TBPN for $100M+ but most people are drawing the wrong conclusions. OpenAI doesn’t care about TBPN as a business or a show. They immediately shut down all ads and they are going to leave full editorial independence to John and Jordi. That is a big win for TBPN and the team, but more importantly it is the signal for what OpenAI was actually after. They wanted to acquire @johncoogan, @jordihays, @DylanAbruscato and the team. These guys are some of the smartest, most experienced marketers on the internet right now. They understand how to win the vibe war, while making everyone love them. Luxury positioning with down-to-earth approachability. Those who have never built a media company won’t understand how difficult it is to pull this off. But John, Jordi, Dylan and others executed this strategy perfectly. And this is what OpenAI has to replicate. They need users to pay for the high compute costs, yet they need to make AI less scary and threatening. So OpenAI is not buying a show. They are not paying a multiple on revenue. And they definitely are not planning to use TBPN to run some psyop campaign. OpenAI just pulled off one of the best talent acquisitions in a long time. They figured out how to hire generational talent to beef up their marketing and comms efforts. The money was merely the price it would take for the team to give up their neutral position in the ecosystem and point their talent behind a single company. OpenAI wanted the people. TBPN was just the vessel to effectuate the transaction. And the money was the market clearing price to pull off a blockbuster trade for first round draft pick talent. Congratulations to each of them. They are true trailblazers who understand their market value and they exploited it to the tune of $100M+ Not bad for a few guys in a warehouse with a couple of cameras :)

@lc_fund @TheLAPurchaser Email harvest pdfs and trade desk blotter. Train your Claude model on how sell side models (download from portal)

@TheLAPurchaser You’d have to use playwright to scrape each portal individually + email ingestion. Claude code isn’t the best at this yet and req some tinkering from my end

What is the best way to tap into sellside research for agentic due diligence? I don’t want any more deep research based on seeking alpha and the motley fool. How do I tap into MS, GS, BAML to create a primer, without plugging in via API?

@KobeissiLetter what navy?

BREAKING: The IRGC says that the Strait of Hormuz is closed and any transit through the waterway will face “harsh measures.” Three container ships of various nationalities were just turned back from the Strait of Hormuz after warnings from the IRGC navy, media reports.

@altryne i run opus 4.6 medium on claude code...make I ever hit in current session before was 30%. I'm even using codex for auditing code so my usage should be lower...i dont get it

@Charter_Digital Dang, and before? was ok?

@altryne i'm not a power user by any means and almost hit it

My feed is showing me a bunch of folks who tapped out their whole usage limits on Mon/Tue. Is this your experience? Please comment, I want to understand how widespread this is

Too many bangers in here, but here are some of my other favorite quotes from this episode 🙏... "I got to spend the last 8.5 years of my career on the Street working at Point72 for Steve Cohen…The competitive advantage at Point72 was about understanding when stories change by one degree instead of ten." "Every piece of content that you share with a PM has to create lift or there is no value in that content…at the end of the day, every PM on the Street is being evaluated on three things. Number of at bats, hit rate against the number of at bats, and then sizing against that hit rate. If you are building a function to inform a PM’s process…you need to help them generate more ideas, have a higher hit rate against their ideas, or help them improve their slugging percentage and/or conviction. If you are not creating lift in one of those three buckets for a hedge fund when you are running the research business, then you don’t have a research business." "In the content business you want to be relevant, different, and accessible." "Alpha is a construct that evolves through time…Alpha moves around. It’s always there, but sometimes it’s in speed to information. Sometimes it’s in just the information itself or access. Sometimes it’s in the organizational ability to process something and turn it into a trade." "Don’t be too married to a view. You want to get it right, not be right. At the end of the day, alpha rewards those that value assets in a cold way…Larry [Robbins] and Steve [Cohen] just wanted to get it right. They were maniacal about getting it right through best practices and best processes." "Having a competitive advantage takes work, and it’s not asking a chatbot." "I used to do thousands of calls talking to CEOs and CFOs of private companies and one of the questions I always ask was ‘when was the last time you saw this? And then what happened after that? And what was the timeline of that?' And then I had a playbook, and I was just matching against the playbook. Once you start thinking in frameworks, stuff really scales." "I would have people on Wall Street learn the old fashioned way [without LLMs] for the first six or twelve months…I sound like an old man, but let’s walk into the room rather than run…I’m a believer in the tools but I also think it’s stunting the growth of this generation…it could lead to degradation in your 30s that you won’t be able to come back from. If you don’t know how to do anything, then you don’t know how to do anything, and competing on your raw smarts isn’t enough because everyone is smart."

@FrankieIsLost Sounds like crypto dev grants

many of the ways in which people use agents today will be completely unviable once the subsidies stop. so might as well make the most out of this brief period where someone else is willing to pick up your inference tab.

Dexter now uses real-time market data. This includes: • stock prices • key ratios • breaking news It’s finally connected to the stock market.

Data centers buying power companies. It’s happening…

Bloomberg trying to clap back at terminal disruption meme's

@ericjackson Watch production customers as a percentage of total deals next quarter...sell side analysts currently buying into the hand waiving

Enterprise SaaS isn’t in trouble because of AI. It’s in trouble if customers don’t deploy what they buy. 29,000 deals. 4,000 live. That’s the real risk.

Ego is the enemy.

@Mike_Scully_ Goldman is building out their own internal LLM that is fed from anthropic API and others. They will soon release, if not already have released a more fine tuned suite of products which run off their internal data sets. NO way is Chatgpt is allowed at work.

I have a close friend who's an investment banker at Goldman Sachs. Trinity graduate. Six figures before bonus. One of the sharpest people I know. Last week I asked him what his day looks like. "Honestly? PowerPoint and Excel. That's probably 80% of my job." Then I asked what AI tools he uses. "ChatGPT sometimes." Have you heard of Claude? "No. What is it?" Let that sink in for a second. A Goldman Sachs investment banker, whose entire job is being faster and smarter than the person next to him, has never heard of the tool that just shipped dedicated plugins for his exact workflow. Investment banking. Equity research. Private equity. Financial analysis. All of it. Built in. Available now. And he doesn't know it exists. But you do. You're not at Goldman. You don't have the degree. You probably don't have his salary. But right now, today, you are ahead of him. You know the tools. You understand what they can do. And more importantly, you know that there are thousands of people exactly like him in every industry. Smart, well paid, completely in the dark. That's not a threat. That's a market. The people making serious money right now aren't the ones with the best credentials. They're the ones who spotted the gap first and started selling into it. You already spotted it. Now do something with it.

Ran through the GitHub for Claude Investment Banker: github.com/anthropics/fin…, Timeline Estimate from Claude but I think we move way faster as IBs see cost savings... Phase: Augmentation Timeframe: Now - 2027 Impact: Analysts use these tools, work faster, same headcount ──────────────────────────────────────── Phase: Compression Timeframe: 2027 - 2029 Impact: Banks realize they need fewer juniors, hiring slows ──────────────────────────────────────── Phase: Restructuring Timeframe: 2029+ Impact: Analyst classes shrink 40-60%, associate role fundamentally changes The data connector ecosystem (S&P, FactSet, PitchBook all building MCP servers) is the real signal. When the data pipes are native, the human-in-the-loop for data gathering vanishes entirely. Bottom line: This doesn't kill investment banking. It kills the apprenticeship model that IB was built on — learn by doing grunt work. The question banks haven't answered yet is: how do you train the next generation of MDs if juniors never manually build a comp?