Offshore
Photo
Bourbon Capital
Seth Klarman: I think that value investing is such a perfect discipline for not blowing up
When you think about Warren Buffet... one thing he really did, Amazingly is never blow up, and never really go through even a rough period and it was some combination of having a structure from the capital being permanent and having the discipline in what he's doing and the combination of that made him relatively unique
I don't think there'll ever be another Buffett...... Such a powerful words right there
tweet
Seth Klarman: I think that value investing is such a perfect discipline for not blowing up
When you think about Warren Buffet... one thing he really did, Amazingly is never blow up, and never really go through even a rough period and it was some combination of having a structure from the capital being permanent and having the discipline in what he's doing and the combination of that made him relatively unique
I don't think there'll ever be another Buffett...... Such a powerful words right there
Seth Klarman Baupost Group Q4 2025
They bought 2,121,391 $AMZN shares https://t.co/uwgBT9A48Z tweet
AkhenOsiris
Morgan Stanley's Nowak on $AMZN
AWS growth can be 30% for quite some time
The second catalyst is agentic commerce. Nowak said Amazon’s growing last-mile inventory, expanding infrastructure and technology investments position it to lead in both vertical and horizontal agentic shopping.
The company’s platform-specific agent Rufus is already contributing 140 basis points to fourth-quarter 2025 gross merchandise value (GMV) growth, he notes.
Amazon has acknowledged the need to “collectively figure out a better customer experience” with horizontal AI agents, and added that “we continue to have a number of conversations,” signaling potential partnerships ahead.
“We look for AMZN horizontal agentic partnerships to emerge, which will make investors feel more confident in AMZN’s long-term positioning,” Nowak wrote.
tweet
Morgan Stanley's Nowak on $AMZN
AWS growth can be 30% for quite some time
The second catalyst is agentic commerce. Nowak said Amazon’s growing last-mile inventory, expanding infrastructure and technology investments position it to lead in both vertical and horizontal agentic shopping.
The company’s platform-specific agent Rufus is already contributing 140 basis points to fourth-quarter 2025 gross merchandise value (GMV) growth, he notes.
Amazon has acknowledged the need to “collectively figure out a better customer experience” with horizontal AI agents, and added that “we continue to have a number of conversations,” signaling potential partnerships ahead.
“We look for AMZN horizontal agentic partnerships to emerge, which will make investors feel more confident in AMZN’s long-term positioning,” Nowak wrote.
tweet
Offshore
Photo
Wasteland Capital
95% chance that the S&P500 just simply goes back up to the top of this bouncy range as we see earnings-driven buying and short covering, and some BTFDOMO on top, in many of the Saaspocalypse victims. https://t.co/ZwZtuPkwkC
tweet
95% chance that the S&P500 just simply goes back up to the top of this bouncy range as we see earnings-driven buying and short covering, and some BTFDOMO on top, in many of the Saaspocalypse victims. https://t.co/ZwZtuPkwkC
tweet
Offshore
Video
Brady Long
companies are gonna post record profits while eliminating the only job available in some towns
tweet
companies are gonna post record profits while eliminating the only job available in some towns
PolyAI has raised $200M from Nvidia, Khosla Ventures, and multiple top VCs.
We're one of the fastest-growing companies in the UK, and we handle 500M+ calls for:
• Marriott
• PG&E
• Gordon Ramsay's restaurants
• And 3,000 more real deployments
Which means that if you've ever called them, chances are you've talked to our voice agents.
Every restaurant we onboard books thousands in revenue within 30 days.
But how?
Because PolyAI works 24/7, answering every call in <2tweet
Offshore
Photo
Javier Blas
U.S. shale is the gift that keeps giving:
“I've been wrong," said Kaes Van't Hof, chief executive of Permian producer Diamondback Energy. "I thought we'd be down by now."
https://t.co/e3JCHMqxMz
tweet
U.S. shale is the gift that keeps giving:
“I've been wrong," said Kaes Van't Hof, chief executive of Permian producer Diamondback Energy. "I thought we'd be down by now."
https://t.co/e3JCHMqxMz
tweet
Moon Dev
Building the Ultimate Solana Sniper: How to Filter 1,000x Gems Before They Ever Hit Birdeye
the ultimate beginners guide to solana trading bots is actually a blueprint for financial freedom because the market is a casino where the house always wins unless you are the one running the code. there is a specific way to see every single token launched on solana before it even reaches a website like birdeye and i am going to show you how to build that filter. i believe that code is the great equalizer because i used to spend hundreds of thousands on developers thinking i could not do it myself but now i live on the edge of the blockchain
the secret to finding 1000x tokens before they hit the trending page starts with the birdeye api which is the most important tool in our arsenal. most people look for gems by staring at a screen all day drinking coffee and hoping for luck but we use computers to do the heavy lifting for us. getting these contract addresses straight from the rpc is the first step but the real magic happens in the filtering logic that removes ninety nine percent of the trash tokens launched every hour
most people look at volume but there is a secret metric called unique wallets per hour that tells you if the volume is real or just one guy washing his own bags. if the unique wallets do not match the trade frequency you know it is a ghost town and you save your money instead of getting liquidated. we pull all data from the api including the total value locked and the twenty four hour trade count to decide which tokens are actually worth our time
if the unique wallets do not match the trade frequency you know it is a ghost town and you save your money. i once spent hundreds of thousands on developers because i thought i could not code but then i discovered the great equalizer that allows an average joe to beat wall street. when you put real money on the line you realize that you need to learn this yourself because nobody else is going to build a bot with your specific edge built into it
now i build my own systems and the iterates to success happen in minutes instead of months. there is a rug pull indicator hidden in the price change history that shows up exactly ten minutes before the floor collapses. we built a filter that looks at the last two hundred and fifty orders to see if the sell pressure is building up to a dangerous level
when you see a ninety percent drop in a five minute candle it is already too late but our code sees the sell pressure building at seventy percent and pulls us out. it is crazy that wall street would never show you this information but that is why i share everything here for the squad. you have to adapt your strategies based on market cap and participant behavior because the regime models are always changing in crypto
the final piece of the puzzle is the execution through jupiter where we can automate the buy and sell orders while we sleep. this sniper bot is designed to find those tiny micro caps under twenty thousand dollars and get in before the massive pump happens. with the bot submitting orders based on cold hard math you finally stop being the exit liquidity for the big players
i believe that anyone can learn this because i was not a coder before i started this journey and i was scared just like you might be. once you realize that coding is just repetition and logic it becomes much easier than learning a second language like spanish. you can take a little risk to learn how to automate your trading and the worst case scenario is that you end up with a high paying job skill
the filtering system we built takes fifteen thousand tokens and drops them down to the best fifteen shots of the day. it uses regex to clean up the symbol names and pandas to organize the data so we can see exactly which tokens have the most potential. we even integrated functions to check if the minting authority is revoked so we do not get caught in a honey pot trap
risk management is the only thing that keeps you in t[...]
Building the Ultimate Solana Sniper: How to Filter 1,000x Gems Before They Ever Hit Birdeye
the ultimate beginners guide to solana trading bots is actually a blueprint for financial freedom because the market is a casino where the house always wins unless you are the one running the code. there is a specific way to see every single token launched on solana before it even reaches a website like birdeye and i am going to show you how to build that filter. i believe that code is the great equalizer because i used to spend hundreds of thousands on developers thinking i could not do it myself but now i live on the edge of the blockchain
the secret to finding 1000x tokens before they hit the trending page starts with the birdeye api which is the most important tool in our arsenal. most people look for gems by staring at a screen all day drinking coffee and hoping for luck but we use computers to do the heavy lifting for us. getting these contract addresses straight from the rpc is the first step but the real magic happens in the filtering logic that removes ninety nine percent of the trash tokens launched every hour
most people look at volume but there is a secret metric called unique wallets per hour that tells you if the volume is real or just one guy washing his own bags. if the unique wallets do not match the trade frequency you know it is a ghost town and you save your money instead of getting liquidated. we pull all data from the api including the total value locked and the twenty four hour trade count to decide which tokens are actually worth our time
if the unique wallets do not match the trade frequency you know it is a ghost town and you save your money. i once spent hundreds of thousands on developers because i thought i could not code but then i discovered the great equalizer that allows an average joe to beat wall street. when you put real money on the line you realize that you need to learn this yourself because nobody else is going to build a bot with your specific edge built into it
now i build my own systems and the iterates to success happen in minutes instead of months. there is a rug pull indicator hidden in the price change history that shows up exactly ten minutes before the floor collapses. we built a filter that looks at the last two hundred and fifty orders to see if the sell pressure is building up to a dangerous level
when you see a ninety percent drop in a five minute candle it is already too late but our code sees the sell pressure building at seventy percent and pulls us out. it is crazy that wall street would never show you this information but that is why i share everything here for the squad. you have to adapt your strategies based on market cap and participant behavior because the regime models are always changing in crypto
the final piece of the puzzle is the execution through jupiter where we can automate the buy and sell orders while we sleep. this sniper bot is designed to find those tiny micro caps under twenty thousand dollars and get in before the massive pump happens. with the bot submitting orders based on cold hard math you finally stop being the exit liquidity for the big players
i believe that anyone can learn this because i was not a coder before i started this journey and i was scared just like you might be. once you realize that coding is just repetition and logic it becomes much easier than learning a second language like spanish. you can take a little risk to learn how to automate your trading and the worst case scenario is that you end up with a high paying job skill
the filtering system we built takes fifteen thousand tokens and drops them down to the best fifteen shots of the day. it uses regex to clean up the symbol names and pandas to organize the data so we can see exactly which tokens have the most potential. we even integrated functions to check if the minting authority is revoked so we do not get caught in a honey pot trap
risk management is the only thing that keeps you in t[...]
Offshore
Moon Dev Building the Ultimate Solana Sniper: How to Filter 1,000x Gems Before They Ever Hit Birdeye the ultimate beginners guide to solana trading bots is actually a blueprint for financial freedom because the market is a casino where the house always wins…
he game when you are trading these degenerate meme coins. i like to throw ten dollars at a bunch of different tokens and if one hits a thousand percent return it covers all the small losses from the rugs. the bot trades emotionally free which is impossible for a human who is watching their life savings bounce around on a chart
you can use api endpoints to see the last trade time and ensure that the token still has active buyers in the last hour. if a token has zero trades in sixty minutes we drop it immediately because it is already a dead ship. it be like that sometimes in the wild west of solana but as long as we have the tools to filter the noise we will find the signal
building a market maker is the next level after you master the sniper bot because it allows you to create your own volume and profit from the spread. we are constantly coming up with new strategies and back testing them to see if they worked in the past before we ever risk a single cent. this is the path from being a gambler to being a quantitative trader who actually understands the math behind the moves
there is so much opportunity in this space right now that it is almost overwhelming but you have to take action to get separation from the crowd. most traders will just keep chasing pumps and losing money because they are too lazy to learn how to code their own edge. i am going to keep building these crazy bots and showing you every single step because i want us all to win together
tweet
you can use api endpoints to see the last trade time and ensure that the token still has active buyers in the last hour. if a token has zero trades in sixty minutes we drop it immediately because it is already a dead ship. it be like that sometimes in the wild west of solana but as long as we have the tools to filter the noise we will find the signal
building a market maker is the next level after you master the sniper bot because it allows you to create your own volume and profit from the spread. we are constantly coming up with new strategies and back testing them to see if they worked in the past before we ever risk a single cent. this is the path from being a gambler to being a quantitative trader who actually understands the math behind the moves
there is so much opportunity in this space right now that it is almost overwhelming but you have to take action to get separation from the crowd. most traders will just keep chasing pumps and losing money because they are too lazy to learn how to code their own edge. i am going to keep building these crazy bots and showing you every single step because i want us all to win together
tweet
X (formerly Twitter)
Moon Dev (@MoonDevOnYT) on X
Building the Ultimate Solana Sniper: How to Filter 1,000x Gems Before They Ever Hit Birdeye
the ultimate beginners guide to solana trading bots is actually a blueprint for financial freedom because the market is a casino where the house always wins unless…
the ultimate beginners guide to solana trading bots is actually a blueprint for financial freedom because the market is a casino where the house always wins unless…
Offshore
Photo
God of Prompt
RT @godofprompt: 🚨 Holy shit… Stanford just published a paper that questions whether we even need humans to study humans.
The title sounds like a joke:
“This human study did not involve human subjects.”
But it’s dead serious.
The researchers are asking a controversial question:
Can LLM simulations count as behavioral evidence?
Here’s the core idea.
Instead of recruiting thousands of participants, running surveys, and waiting weeks for results, they simulate people using large language models.
Not generic prompts.
But structured simulations where the model is assigned demographic traits, preferences, beliefs, and contextual constraints.
Then they test whether the simulated responses statistically match real-world human data.
And disturbingly… they often do.
Across multiple behavioral tasks, the LLM-generated “participants” reproduced known human patterns:
• Established psychological biases
• Preference distributions
• Decision-making trends
• Even demographic splits
Not perfectly. Not universally.
But far closer than most people would expect.
The key contribution of the paper isn’t “LLMs are human.”
It’s validation.
They systematically compare simulated outputs to ground-truth human datasets and evaluate alignment using statistical benchmarks.
When the distributions match, the simulation isn’t just storytelling.
It becomes empirical evidence.
That’s the uncomfortable shift.
If a sufficiently constrained LLM simulation reproduces real behavioral patterns, does it become a legitimate experimental proxy?
Because if the answer is yes, this changes everything:
• Behavioral economics
• Political science
• Market research
• Policy testing
• UX experimentation
You could prototype social interventions before deploying them in the real world.
You could stress-test messaging strategies across simulated demographics.
You could explore rare edge-case populations without recruitment bottlenecks.
But here’s where Stanford is careful.
The models don’t “understand” humans.
They reflect training data patterns.
They can amplify biases.
They can collapse under distribution shift.
And they can simulate plausibility without causality.
So the paper doesn’t claim replacement.
It argues for calibration.
LLM simulations can be useful behavioral instruments if validated against real data and bounded within known limits.
That’s the distinction.
Not synthetic humans.
Synthetic behavioral priors.
The wild part?
This paper forces academia to confront something bigger:
If large models encode large-scale behavioral regularities from the internet, they become compressed maps of human tendencies.
Not minds.
Maps.
And maps can be useful.
We’re moving from “AI as text generator” to “AI as behavioral simulator.”
The ethics, methodology, and epistemology implications are massive.
Because once simulation becomes statistically reliable, the bottleneck in social science shifts from data collection to model alignment.
And that might be the real revolution hidden in this paper.
tweet
RT @godofprompt: 🚨 Holy shit… Stanford just published a paper that questions whether we even need humans to study humans.
The title sounds like a joke:
“This human study did not involve human subjects.”
But it’s dead serious.
The researchers are asking a controversial question:
Can LLM simulations count as behavioral evidence?
Here’s the core idea.
Instead of recruiting thousands of participants, running surveys, and waiting weeks for results, they simulate people using large language models.
Not generic prompts.
But structured simulations where the model is assigned demographic traits, preferences, beliefs, and contextual constraints.
Then they test whether the simulated responses statistically match real-world human data.
And disturbingly… they often do.
Across multiple behavioral tasks, the LLM-generated “participants” reproduced known human patterns:
• Established psychological biases
• Preference distributions
• Decision-making trends
• Even demographic splits
Not perfectly. Not universally.
But far closer than most people would expect.
The key contribution of the paper isn’t “LLMs are human.”
It’s validation.
They systematically compare simulated outputs to ground-truth human datasets and evaluate alignment using statistical benchmarks.
When the distributions match, the simulation isn’t just storytelling.
It becomes empirical evidence.
That’s the uncomfortable shift.
If a sufficiently constrained LLM simulation reproduces real behavioral patterns, does it become a legitimate experimental proxy?
Because if the answer is yes, this changes everything:
• Behavioral economics
• Political science
• Market research
• Policy testing
• UX experimentation
You could prototype social interventions before deploying them in the real world.
You could stress-test messaging strategies across simulated demographics.
You could explore rare edge-case populations without recruitment bottlenecks.
But here’s where Stanford is careful.
The models don’t “understand” humans.
They reflect training data patterns.
They can amplify biases.
They can collapse under distribution shift.
And they can simulate plausibility without causality.
So the paper doesn’t claim replacement.
It argues for calibration.
LLM simulations can be useful behavioral instruments if validated against real data and bounded within known limits.
That’s the distinction.
Not synthetic humans.
Synthetic behavioral priors.
The wild part?
This paper forces academia to confront something bigger:
If large models encode large-scale behavioral regularities from the internet, they become compressed maps of human tendencies.
Not minds.
Maps.
And maps can be useful.
We’re moving from “AI as text generator” to “AI as behavioral simulator.”
The ethics, methodology, and epistemology implications are massive.
Because once simulation becomes statistically reliable, the bottleneck in social science shifts from data collection to model alignment.
And that might be the real revolution hidden in this paper.
tweet
Offshore
Photo
Benjamin Hernandez😎
$WING 14% EPS COMPOUNDER!
Wingstop +14.55% hitting $288.41. EPS growth is +79.0% YoY. Buy ratings across the board. The fundamental moat is widening.
Buy on any shallow pullback, the trend is unstoppable.
DM to get the specific Wave 3 entry points.
$SOC $BMNR $BYND $NB $PULM https://t.co/zSGxt9jfZ9
tweet
$WING 14% EPS COMPOUNDER!
Wingstop +14.55% hitting $288.41. EPS growth is +79.0% YoY. Buy ratings across the board. The fundamental moat is widening.
Buy on any shallow pullback, the trend is unstoppable.
DM to get the specific Wave 3 entry points.
$SOC $BMNR $BYND $NB $PULM https://t.co/zSGxt9jfZ9
tweet
Offshore
Video
Brady Long
RT @thisdudelikesAI: the last time something this big happened to an industry this fast was kodak. and kodak had a warning...
tweet
RT @thisdudelikesAI: the last time something this big happened to an industry this fast was kodak. and kodak had a warning...
PolyAI has raised $200M from Nvidia, Khosla Ventures, and multiple top VCs.
We're one of the fastest-growing companies in the UK, and we handle 500M+ calls for:
• Marriott
• PG&E
• Gordon Ramsay's restaurants
• And 3,000 more real deployments
Which means that if you've ever called them, chances are you've talked to our voice agents.
Every restaurant we onboard books thousands in revenue within 30 days.
But how?
Because PolyAI works 24/7, answering every call in <2tweet