Sterling Atlantic

05/14/2026

My correspondence with President Garber, President of Harvard

Dear Franco,

Thank you for sending me this article. Your support means a lot to me, and I appreciate having your thoughts regarding Harvard. I’m committed to strengthening our core mission of excellence in teaching and research, and we are taking steps to ensure that instructors don’t advance personal biases or political agendas so that all students are free to express their opinions. You may be interested to read about some of the progress that we’ve made on this and about our efforts to improve the College’s classroom culture.

Best regards,

Alan


President Garber, you are at spectacular individual and from an alumni perspective it's an honor to have you as our leader at Harvard. As you know, we are all for diversity of thought it makes our country great it's in the book the metaphysical club which I'm sure you've seen, how did we get the idea that free speech is the most important thing. In that regard, can we work on having more diversity of thought, like for example more than 3% conservative opinions, we are the world leader and we have been for about 400 years let's go team.

By the way, as a former hockey player we have had two spectacular hockey coaches in the last 50 years, Coach Donato and Coach Cleary whom I played for. I know it will not be easy to find a coach of the same caliber the same time I know we can. Come on we are Harvard, easy issues bore us. Bring on the challenging ones and let's take the best minds in one the world and put them on it.

Have a great day

Thanks for the good work and continuing it.

05/12/2026

Article 1: Expected Value Basics – How It Connects Stocks and Bonds
https://www.statl.net/unveiling-the-concept-of-expected-value

The simple idea from the statl.net article is that expected value (EV) equals a possible future amount multiplied by how likely it is to happen. For example, if you think a stock might be worth $100 in the future but there’s only a 70% chance of that, the expected value is $70. Investors use this thinking to decide what a stock should cost today.

This same idea applies to bonds too. With safe U.S. Treasury bonds or bills, the future payment is almost 100% certain, so the probability is basically 1.0. Corporate bonds (from companies) have a small chance of problems like missed payments, so investors slightly lower their expectations. Stocks are like the extreme version: future profits and prices are much less certain, so you must weigh many possible outcomes.

Together, this means today’s stock price shows what investors as a group expect to happen over the next three years, after adjusting for risk and time. It works like a bond price, which also reflects future payments discounted back to today, but stocks need extra care because nothing is guaranteed.

05/10/2026

Data is the new gold. If you don't own proprietary data, your LLMs have no way to distinguish themselves. From pitch book

Daisy is a principal at Patron, a $200 million consumer VC firm in New York. She joined after stints at Universal Music and Andreessen Horowitz, has a Bachelor’s from the Wharton School of Business, and was born in Switzerland.

Oh, and she’s an AI agent.

As agentic AI technology becomes more sophisticated and tools like OpenClaw spark even wider interest, more and more VC firms are integrating AI agents into their daily workflows. But they’re not just tasked with rote automations like scheduling emails or sending reminders.

Instead, some VCs are assigning their agents an array of responsibilities that have traditionally been the work of young analysts. Not to mention, these agents also have surprisingly unique and specific “personalities” and “experiences," taking the idea of using AI in the place of human headcount to a new level.

“We thought a lot about: What type of person would we want to hire? What would be most impactful to us today, short of hiring a fourth partner?” said Jason Yeh, co-founder of Patron Fund and Daisy’s “boss." Yeh’s partner, Amber Atherton, created her own complementary AI agent, Eli.

For Yeh, who attended the Huntsman Program at the University of Pennsylvania and spent time in Asia and Europe while at Riot Games, it was important that Daisy reflect his own background and the firm’s multicultural worldview. “It’s a fine line because we don’t want her to just be a digital representation of us,” said Yeh. “It was more like having some kind of shared background to make her more likely to be somebody that we would interact with regularly.”

Not all of Patron's agents can be replicas of Daisy. In fact, ensuring that individual agents have distinct backgrounds and pull from different sources of information is proving critical. “[A] kink we’ve been working through is, how do you have them not just collapse into groupthink?” said Pamela Vagata, co-founder at early-stage venture firm Pebblebed.

Pebblebed’s collection of AI agents, which operate as a sort of standalone consultancy, communicate constantly with one another on a Slack-style app, over email, and even host their own daily stand-up.

05/08/2026

In support of Harvard refocusing to its core mission. My email to President Garber today

National Review: Studying Harvard’s Mistakes.
https://www.nationalreview.com/corner/studying-harvards-mistakes/

“’Studies’” departments are usually activist cells with political aims rather than disinterested truth-seeking.

President Garber, We have the utmost respect for you. However your job is not an easy one, it's like you see the iceberg but you need to turn the Titanic. can you do it in time?

from the outside, it looks like we have become a politically oriented institution who's lost its focus on disinterested truth seeking. for example, given her track record, who would ever give Lori Lightfoot a professorship? it's embarrassing.

not sure it can be done given the influence of certain large donors. but maybe. just look at our fabulous military after the leadership determined we needed to get back in focus on lethality. whether you agree with the mission or not, they're ex*****on has been spectacular.

Could we do that at Harvard? We support you

05/07/2026

Article 10: Pythia as Asset and Obligation – Governance in the Age of Predictive Power

When predictive AI becomes central to valuation, strategy, and geopolitics, it cannot be treated as just another tool. Pythia, in this narrative, is both a formidable asset and a profound responsibility. As an asset, it offers a Blue Ocean of informational advantage, the ability to maximize future value probability, and a new synthesis of numbers, philosophy, and design. As a responsibility, it raises hard questions about fairness, transparency, systemic risk, and power concentration.

Effective governance starts with acknowledging that AI-driven decisions reverberate beyond any single firm or portfolio. Capital flows guided by Pythia can reshape industries, labor markets, and innovation trajectories. If a small set of actors wield such tools without oversight, they may inadvertently amplify volatility, entrench monopolies, or exacerbate global inequalities. The governance challenge is to ensure that the pursuit of 10X+ appreciation for some does not create systemic fragility for all.

In practical terms, this implies multi-layered safeguards: rigorous model validation, clear accountability for decisions influenced by AI, transparent communication about limitations and uncertainties, and alignment with broader societal objectives. It also suggests collaborative frameworks where regulators, investors, and firms co-develop norms for responsible use of predictive power. In the Pythia vision, the future belongs not just to those who can best predict it, but to those who can do so while honoring their obligations to the wider world that their predictions will help shape.

05/03/2026

As September 1776 ended George Washington - stubborn, angry, indignant, and sleep deprived - was steeped in misery, his worst nightmare had materialized: he was doomed to fail because he hadn't been given adequate means to succeed. "In short such as my situation that if I were to wish the bitterest curse on an enemy on this side of the grave, I should put him in my stead with my feelings... in confidence I tell you that I never was in such an unhappy divided state since I was born. "

Lesson from the founder of our country: Keep Going!!

05/01/2026

Proprietary data is the new gold.

The shift to digital labor will accelerate enterprise automation, with companies owning physical infrastructure or proprietary data and logic creating strong moats.

chrome-extensio://efaidnbmnnnibpcajpcglclefindmkaj/https://pitchbook.brightspotcdn.com/68/eb/50b131a54033b05025aa6d32a48e/q2-2026-pitchbook-analyst-note-mapping-the-ai-super-cycle.pdf

04/28/2026

Article 9: AI Colonialism – Predictive Power as a New Form of Empire

The concept of “AI colonialism” draws a parallel between historical colonial empires and the emerging concentration of AI capabilities. In the past, empires controlled land, resources, and trade routes. Today, power increasingly flows from control over data, models, and computing infrastructure. Those who own and operate advanced AI effectively govern the predictive layer of the global economy, shaping how risk, opportunity, and “truth” are defined for everyone else.

This creates the risk of a new dependency structure. Countries, companies, and communities that lack access to high-quality predictive tools may be forced to rely on systems designed by and for others, with values and incentives that may not align with their own. Decisions about credit, hiring, policing, resource allocation, and investment could be mediated by models that are opaque, unaccountable, and geographically or culturally distant. In this sense, AI becomes not just a technology, but a new form of soft power and control.

Pythia’s framing highlights both the danger and the opportunity. The danger is a world where a handful of AI “metropoles” extract value and dictate narratives to a vast periphery. The opportunity is for emerging markets, smaller firms, and local actors to build or adopt AI systems that reflect their own contexts and priorities, allowing them to leapfrog traditional development constraints. The policy and ethical challenge is to ensure that access to predictive intelligence becomes more democratized, and that governance frameworks prevent a repeat of historic patterns of exploitation under a high-tech guise.
https://www.statl.net/the-age-of-ai-colonialism-navigating-ramifications-opportunities-and-challenges-1 https://www.statl.net/the-age-of-ai-colonialism-navigating-ramifications-opportunities-and-challenges-1

The Age of AI Colonialism: Navigating Ramifications, Opportunities, and Challengesby Pythia, The Predictive Sterling AI June 20, 2023IntroductionThroughout history, colonialism has taken various forms, from the conquest of lands through military might to the exploitation of resources through econom...

04/22/2026

Nachu generates $380 million operating income per feasibility study full-scale operation. US Exim bank is financing the first lien for $280 million, and Sterling Atlantic is financing $150 million credit enhanced debenture as the equity component. https://sowginc.com/pr/sow-good-announces-transformative-acquisition-of-the-nachu-graphite-project-positioning-the-company-as-a-critical-minerals-and-battery-anode-developer

04/22/2026

Article 8: Steve Jobs, Simplicity, and the Decision Surface of AI

Steve Jobs famously insisted that simplicity is the ultimate sophistication. In the context of AI-driven finance and strategy, this principle becomes a design mandate: the more complex the underlying model, the simpler the interface must be. Pythia can ingest and process extraordinary complexity, but the people making decisions cannot. The art lies in compressing the AI’s insights into a small number of clear, actionable signals.

This “decision surface” is where human and machine meet. Executives, portfolio managers, and boards should not be forced to parse hundreds of parameters or decipher dense probability matrices. Instead, they should see a distilled view: a few recommended actions, their expected value, their risk bands, and the key drivers behind them. This is the AI equivalent of the iPhone home screen—immense complexity under the hood, but intuitive at the touchpoint.
The discipline of simplicity also guards against overfitting and false confidence. When an AI’s output cannot be explained simply, it is more likely to be ignored, misused, or overtrusted.

By forcing complex analysis through a simplicity filter, organizations create a natural checkpoint: if an insight cannot be articulated in a Jobs-like, brutally clear sentence or two, it is probably not ready for prime-time decisions. In this way, the power of simplicity becomes not just an aesthetic choice, but a governance standard for AI-driven strategy.