Offshore
Photo
Dimitry Nakhla | Babylon Capitalยฎ
Moodyโs $MCO Q4 25โ Report๐๏ธ
โ REV: $1.89B (+13% YoY)
โ EPS: $3.64 (+39% YoY) https://t.co/CNJFpAduw1
tweet
Moodyโs $MCO Q4 25โ Report๐๏ธ
โ REV: $1.89B (+13% YoY)
โ EPS: $3.64 (+39% YoY) https://t.co/CNJFpAduw1
tweet
Offshore
Photo
God of Prompt
RT @godofprompt: ๐จ Holy shitโฆ Stanford just published a paper that questions whether we even need humans to study humans.
The title sounds like a joke:
โThis human study did not involve human subjects.โ
But itโs dead serious.
The researchers are asking a controversial question:
Can LLM simulations count as behavioral evidence?
Hereโs 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 disturbinglyโฆ they often do.
Across multiple behavioral tasks, the LLM-generated โparticipantsโ reproduced known human patterns:
โข Established psychological biases
โข Preference distributions
โข Decision-making trends
โข Even demographic splits
Not perfectly. Not universally.
But far closer than most people would expect.
The key contribution of the paper isnโt โLLMs are human.โ
Itโs validation.
They systematically compare simulated outputs to ground-truth human datasets and evaluate alignment using statistical benchmarks.
When the distributions match, the simulation isnโt just storytelling.
It becomes empirical evidence.
Thatโs 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:
โข Behavioral economics
โข Political science
โข Market research
โข Policy testing
โข 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 hereโs where Stanford is careful.
The models donโt โunderstandโ 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 doesnโt claim replacement.
It argues for calibration.
LLM simulations can be useful behavioral instruments if validated against real data and bounded within known limits.
Thatโs 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.
Weโre moving from โAI as text generatorโ to โAI as behavioral simulator.โ
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: ๐จ Holy shitโฆ Stanford just published a paper that questions whether we even need humans to study humans.
The title sounds like a joke:
โThis human study did not involve human subjects.โ
But itโs dead serious.
The researchers are asking a controversial question:
Can LLM simulations count as behavioral evidence?
Hereโs 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 disturbinglyโฆ they often do.
Across multiple behavioral tasks, the LLM-generated โparticipantsโ reproduced known human patterns:
โข Established psychological biases
โข Preference distributions
โข Decision-making trends
โข Even demographic splits
Not perfectly. Not universally.
But far closer than most people would expect.
The key contribution of the paper isnโt โLLMs are human.โ
Itโs validation.
They systematically compare simulated outputs to ground-truth human datasets and evaluate alignment using statistical benchmarks.
When the distributions match, the simulation isnโt just storytelling.
It becomes empirical evidence.
Thatโs 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:
โข Behavioral economics
โข Political science
โข Market research
โข Policy testing
โข 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 hereโs where Stanford is careful.
The models donโt โunderstandโ 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 doesnโt claim replacement.
It argues for calibration.
LLM simulations can be useful behavioral instruments if validated against real data and bounded within known limits.
Thatโs 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.
Weโre moving from โAI as text generatorโ to โAI as behavioral simulator.โ
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
DAIR.AI
RT @omarsar0: // From Vibe Coding to Agentic Engineering //
GLM-5 is a foundation model designed to transition from vibe coding to agentic engineering.
The model introduces novel asynchronous agent RL algorithms that enable learning from complex, long-horizon interactions. It also adopts DSA to reduce computational costs while preserving long-context performance.
The key contribution is an asynchronous RL infrastructure that decouples generation from training, allowing the model to learn from extended agentic workflows rather than short isolated tasks.
GLM-5 demonstrates strong performance on standard benchmarks and surpasses previous baselines in handling end-to-end software engineering challenges.
Paper: https://t.co/pl50bRSXVR
Learn to build effective AI agents in our academy: https://t.co/1e8RZKs4uX
tweet
RT @omarsar0: // From Vibe Coding to Agentic Engineering //
GLM-5 is a foundation model designed to transition from vibe coding to agentic engineering.
The model introduces novel asynchronous agent RL algorithms that enable learning from complex, long-horizon interactions. It also adopts DSA to reduce computational costs while preserving long-context performance.
The key contribution is an asynchronous RL infrastructure that decouples generation from training, allowing the model to learn from extended agentic workflows rather than short isolated tasks.
GLM-5 demonstrates strong performance on standard benchmarks and surpasses previous baselines in handling end-to-end software engineering challenges.
Paper: https://t.co/pl50bRSXVR
Learn to build effective AI agents in our academy: https://t.co/1e8RZKs4uX
tweet
Offshore
Photo
DAIR.AI
RT @dair_ai: A paper worth paying close attention to.
It presents Lossless Context Management (LCM), which reframes how agents handle long contexts.
It outperforms Claude Code on long-context tasks.
Recursive Language Models give the model full autonomy to write its own memory scripts. LCM takes that power back, handing it to a deterministic engine that compresses old messages into a hierarchical DAG while keeping lossless pointers to every original. Less expressive in theory, far more reliable in practice.
The results:
Their agent (Volt, on Opus 4.6) beats Claude Code at *every* context length from 32K to 1M tokens on the OOLONG benchmark. +29.2 points average improvement versus Claude Code's +24.7. The gap widens at longer contexts.
The implication is one we keep relearning from software engineering history: how you manage what the model sees may matter more than giving the model tools to manage it itself. Every agent framework shipping with "let the model figure it out" memory strategies may be building on the wrong abstraction entirely.
Paper: https://t.co/LtqS7pzmP4
Learn to build effective AI agents in our academy: https://t.co/LRnpZN7L4c
tweet
RT @dair_ai: A paper worth paying close attention to.
It presents Lossless Context Management (LCM), which reframes how agents handle long contexts.
It outperforms Claude Code on long-context tasks.
Recursive Language Models give the model full autonomy to write its own memory scripts. LCM takes that power back, handing it to a deterministic engine that compresses old messages into a hierarchical DAG while keeping lossless pointers to every original. Less expressive in theory, far more reliable in practice.
The results:
Their agent (Volt, on Opus 4.6) beats Claude Code at *every* context length from 32K to 1M tokens on the OOLONG benchmark. +29.2 points average improvement versus Claude Code's +24.7. The gap widens at longer contexts.
The implication is one we keep relearning from software engineering history: how you manage what the model sees may matter more than giving the model tools to manage it itself. Every agent framework shipping with "let the model figure it out" memory strategies may be building on the wrong abstraction entirely.
Paper: https://t.co/LtqS7pzmP4
Learn to build effective AI agents in our academy: https://t.co/LRnpZN7L4c
tweet
Offshore
Photo
Dimitry Nakhla | Babylon Capitalยฎ
RT @patrick_oshag: This is my second conversation with @JoshuaKushner.
Josh started Thrive in 2011 and the firm now manages ~$50 billion. We cover the iconic investments that defined it: Instagram, Stripe, GitHub, and spend a lot of time on OpenAI. He explains how Thrive thinks about investing today and the three categories they're currently focused on.
Josh also talks about how he built the firm โ why they keep the team so small, why concentration is core to what they do, and what he's learned from A24 about enabling artists to create their best work.
Throughout the conversation, Josh shares the personal stories that shaped him, from his grandmother surviving the Holocaust to lessons from Stan Druckenmiller and Jon Winkelried at formative moments in Thrive's history.
Enjoy!
https://t.co/B0ZMk6Oydo
tweet
RT @patrick_oshag: This is my second conversation with @JoshuaKushner.
Josh started Thrive in 2011 and the firm now manages ~$50 billion. We cover the iconic investments that defined it: Instagram, Stripe, GitHub, and spend a lot of time on OpenAI. He explains how Thrive thinks about investing today and the three categories they're currently focused on.
Josh also talks about how he built the firm โ why they keep the team so small, why concentration is core to what they do, and what he's learned from A24 about enabling artists to create their best work.
Throughout the conversation, Josh shares the personal stories that shaped him, from his grandmother surviving the Holocaust to lessons from Stan Druckenmiller and Jon Winkelried at formative moments in Thrive's history.
Enjoy!
https://t.co/B0ZMk6Oydo
tweet
Offshore
Photo
Dimitry Nakhla | Babylon Capitalยฎ
RT @DimitryNakhla: Daniel Loeb has bought & completely sold $META four separate times over the past decade.
๐๐๐ ๐๐ ๐จ๐๐ข๐ฅ๐ก๐ฎ ๐๐๐ก๐ his original 3.75M shares, that stake would be worth roughly $2.40B today โ nearly 1/3 of Third Pointโs reported current assets (latest Q4 โ25 13F).
___
No โ this is not a knock on Loeb.
Heโs far smarter & more successful than me.
๐๐ฎ๐ญ ๐ญ๐ก๐๐ซ๐โ๐ฌ ๐๐ง ๐ข๐ฆ๐ฉ๐จ๐ซ๐ญ๐๐ง๐ญ ๐ฅ๐๐ฌ๐ฌ๐จ๐ง ๐ก๐๐ซ๐:
For those of us fortunate enough to own a truly exceptional business, the hardest (yet often most profitable) strategy can be:
๐๐จ๐ข๐ง๐ ๐ง๐จ๐ญ๐ก๐ข๐ง๐ .
As Charlie Munger said:
โ๐๐ฏ๐ท๐ฆ๐ด๐ต๐ช๐ฏ๐จ ๐ช๐ด ๐ธ๐ฉ๐ฆ๐ณ๐ฆ ๐บ๐ฐ๐ถ ๐ง๐ช๐ฏ๐ฅ ๐ข ๐ง๐ฆ๐ธ ๐จ๐ณ๐ฆ๐ข๐ต ๐ค๐ฐ๐ฎ๐ฑ๐ข๐ฏ๐ช๐ฆ๐ด ๐ข๐ฏ๐ฅ ๐ต๐ฉ๐ฆ๐ฏ ๐ด๐ช๐ต ๐ฐ๐ฏ ๐บ๐ฐ๐ถ๐ณ ๐ข๐ด๐ด.โ
tweet
RT @DimitryNakhla: Daniel Loeb has bought & completely sold $META four separate times over the past decade.
๐๐๐ ๐๐ ๐จ๐๐ข๐ฅ๐ก๐ฎ ๐๐๐ก๐ his original 3.75M shares, that stake would be worth roughly $2.40B today โ nearly 1/3 of Third Pointโs reported current assets (latest Q4 โ25 13F).
___
No โ this is not a knock on Loeb.
Heโs far smarter & more successful than me.
๐๐ฎ๐ญ ๐ญ๐ก๐๐ซ๐โ๐ฌ ๐๐ง ๐ข๐ฆ๐ฉ๐จ๐ซ๐ญ๐๐ง๐ญ ๐ฅ๐๐ฌ๐ฌ๐จ๐ง ๐ก๐๐ซ๐:
For those of us fortunate enough to own a truly exceptional business, the hardest (yet often most profitable) strategy can be:
๐๐จ๐ข๐ง๐ ๐ง๐จ๐ญ๐ก๐ข๐ง๐ .
As Charlie Munger said:
โ๐๐ฏ๐ท๐ฆ๐ด๐ต๐ช๐ฏ๐จ ๐ช๐ด ๐ธ๐ฉ๐ฆ๐ณ๐ฆ ๐บ๐ฐ๐ถ ๐ง๐ช๐ฏ๐ฅ ๐ข ๐ง๐ฆ๐ธ ๐จ๐ณ๐ฆ๐ข๐ต ๐ค๐ฐ๐ฎ๐ฑ๐ข๐ฏ๐ช๐ฆ๐ด ๐ข๐ฏ๐ฅ ๐ต๐ฉ๐ฆ๐ฏ ๐ด๐ช๐ต ๐ฐ๐ฏ ๐บ๐ฐ๐ถ๐ณ ๐ข๐ด๐ด.โ
tweet
Offshore
Photo
Illiquid
So far weโve gotten stories about Towa, Nittobo, Fujibo, and Toto. Itโs just a matter of time before local press see the traffic and follow the playbook. Every Asian country has a compelling AI supplier story.
tweet
So far weโve gotten stories about Towa, Nittobo, Fujibo, and Toto. Itโs just a matter of time before local press see the traffic and follow the playbook. Every Asian country has a compelling AI supplier story.
A very 2026 headline. https://t.co/kSZWtjD3yC tweet
Jukan
Some of you have been asking me for my thoughts on Raspberry Pi stock, so let me be honest.
I used to trade meme stocks myself. The result was a significant loss, and I don't touch them anymore.
There's one lesson I learned from that experience โ and it was an expensive one:
If you're going to get swept up in the hype, get swept up early.
Meme stocks are ultimately a game of narrative and momentum. Either you get in early and take profits, or you don't touch it at all. The worst-case scenario is getting swept up late and entering after the move has already been made. By that point, you're just someone else's exit liquidity.
I have no opinion on Raspberry Pi. I won't be touching it. These kinds of names are nearly impossible to predict or model on a fundamental basis โ how much they'll go up, how much they'll come down. I don't bet in areas where I have no edge.
tweet
Some of you have been asking me for my thoughts on Raspberry Pi stock, so let me be honest.
I used to trade meme stocks myself. The result was a significant loss, and I don't touch them anymore.
There's one lesson I learned from that experience โ and it was an expensive one:
If you're going to get swept up in the hype, get swept up early.
Meme stocks are ultimately a game of narrative and momentum. Either you get in early and take profits, or you don't touch it at all. The worst-case scenario is getting swept up late and entering after the move has already been made. By that point, you're just someone else's exit liquidity.
I have no opinion on Raspberry Pi. I won't be touching it. These kinds of names are nearly impossible to predict or model on a fundamental basis โ how much they'll go up, how much they'll come down. I don't bet in areas where I have no edge.
tweet
Javier Blas
RT @LindseyGrahamSC: To those who are perpetuating false narratives against the United Arab Emirates and President Sheikh @MohamedBinZayed personally, you are full of it. I met with him today for an hour and a half. Not only is he alive, but he is also well and as sharp as Iโve ever seen him. To those powers that feel the need to attack MbZ and the UAE for doing the right thing - you do so at your own peril.
Our meeting today was very enjoyable and informative. We discussed the historic moment that is facing the region. I told him how much I appreciated his courage and vision to create an Islamic country that can be integrated into the world in a win-win fashion, both for the people of the UAE and for those who visit and do business with the country.
However, there are other voices in Islam that have the darkest vision of mankind. Those voices are distinctly in the minority, in my view.
MbZโs decision to embrace the Abraham Accords and to modernize his country while still maintaining the faith is the biggest change in the Middle East in my lifetime. What the United Arab Emirates have done to try to integrate the region with the whole world is one of the bravest and most consequential decisions any Middle Eastern leader has made. I was very candid with MbZ that he cannot do this by himself. Other people in the region have to buy-in to whatโs happening with the UAE, not just be casual observers.
To the region: Understand that history is about to be made. President Trump wants a region that looks more like the UAE and less like the Ayatollah. The region can only move forward if it follows the vision that embraces the light instead of going backwards into the darkness. The UAEโs vision for the Middle East and the 2030 vision previously expressed by the Crown Prince of Saudi Arabia is something I would fully embrace because it would be great for South Carolina, and great for America.
The forces that are merging here recently are trying to undercut the movement toward the light. They are going back to the old way of doing business, playing cheap politics. Your actions have not gone unnoticed by me or others. If this continues, it will do enormous damage to the best opportunity I've seen in hundreds of years to change the Middle East for the better.
Finally, to those who believe that the region still flourishes if the ayatollahโs regime survives, I could not disagree more. If this religious Nazi regime in Iran still stands after all this bluster and the people are shut out and continue to be oppressed, it puts everything weโve worked for at risk, including the Abraham Accords.
Now, I am off to Saudi Arabia where I look forward to meeting with the Crown Prince who has shown a lot of courage and wisdom and has embraced, in the past, a vision that will forever change the Middle East for the better.
Time will tell as to what happens.
tweet
RT @LindseyGrahamSC: To those who are perpetuating false narratives against the United Arab Emirates and President Sheikh @MohamedBinZayed personally, you are full of it. I met with him today for an hour and a half. Not only is he alive, but he is also well and as sharp as Iโve ever seen him. To those powers that feel the need to attack MbZ and the UAE for doing the right thing - you do so at your own peril.
Our meeting today was very enjoyable and informative. We discussed the historic moment that is facing the region. I told him how much I appreciated his courage and vision to create an Islamic country that can be integrated into the world in a win-win fashion, both for the people of the UAE and for those who visit and do business with the country.
However, there are other voices in Islam that have the darkest vision of mankind. Those voices are distinctly in the minority, in my view.
MbZโs decision to embrace the Abraham Accords and to modernize his country while still maintaining the faith is the biggest change in the Middle East in my lifetime. What the United Arab Emirates have done to try to integrate the region with the whole world is one of the bravest and most consequential decisions any Middle Eastern leader has made. I was very candid with MbZ that he cannot do this by himself. Other people in the region have to buy-in to whatโs happening with the UAE, not just be casual observers.
To the region: Understand that history is about to be made. President Trump wants a region that looks more like the UAE and less like the Ayatollah. The region can only move forward if it follows the vision that embraces the light instead of going backwards into the darkness. The UAEโs vision for the Middle East and the 2030 vision previously expressed by the Crown Prince of Saudi Arabia is something I would fully embrace because it would be great for South Carolina, and great for America.
The forces that are merging here recently are trying to undercut the movement toward the light. They are going back to the old way of doing business, playing cheap politics. Your actions have not gone unnoticed by me or others. If this continues, it will do enormous damage to the best opportunity I've seen in hundreds of years to change the Middle East for the better.
Finally, to those who believe that the region still flourishes if the ayatollahโs regime survives, I could not disagree more. If this religious Nazi regime in Iran still stands after all this bluster and the people are shut out and continue to be oppressed, it puts everything weโve worked for at risk, including the Abraham Accords.
Now, I am off to Saudi Arabia where I look forward to meeting with the Crown Prince who has shown a lot of courage and wisdom and has embraced, in the past, a vision that will forever change the Middle East for the better.
Time will tell as to what happens.
tweet
Offshore
Photo
Chips & SaaS
RT @Noahpinion: The China Cycle came for Tesla in the end, as everyone with half a brain knew it would. https://t.co/TK1GZ6IVw8
tweet
RT @Noahpinion: The China Cycle came for Tesla in the end, as everyone with half a brain knew it would. https://t.co/TK1GZ6IVw8
Chinese state media amplifies viral Tesla failure amid 45% sales crash https://t.co/OpF53hMJMw by @fredlambert tweet