Chamath Palihapitiya Channel

Prior to the IPO, my team assembled a Deep Dive to understand the business model of SpaceX from first principles. Their latest Q3 results reveal a significant shift, as the Super Intelligence business officially surpassed the combined revenue of the Space and Connectivity businesses. Driven largely by new deals focused on renting out computing capacity to other companies, Super Intelligence brought in $7.6B. This figure represents ~52% of the total revenue for the quarter, marking a substantial leap from just 15% a year ago. Furthermore, the company reported an overall capex of $19.6B during the quarter, an amount that actually exceeds its entire revenue. Of those capital expenditures, a sizable $16.4B was allocated directly to Super Intelligence.

Read it here: https://research.socialcapital.com/go/spacex/x/202610231700-chart-post/chamath

While I receive pitches from entrepreneurs on a daily basis, I want to invite you to a much more effective alternative: joining our hackathon.

The event is called EVOLVE. It will span an overnight duration of 20 hours from Nov 13 to Nov 14, hosted at 8090 in Redwood City. Your objective will be to construct AI agents and tools that are capable of building software.

Along with a great trophy and public recognition, the champions will earn two 1:1 meetings with me. The first session is designed to help you polish an idea or pitch, and the second will serve as a follow-up. If you secure the win, you have total control over how we spend that time together. You are fully welcome to request money, seek my advice, or simply spend the time making fun of me.

Submit your application by visiting: https://8090-solutions.link/evolve-with-chamath

The current data completely aligns with my own viewing habits. While my Netflix subscription essentially sits unused, I find myself streaming videos on YouTube on a daily basis.

From a business perspective, the financial contrast between the two platforms is quite striking. YouTube benefits from incredibly low expenses when it comes to content production. Conversely, Netflix currently carries $14B in debt, which was specifically utilized to build up its extensive media library.

Could this situation ultimately serve as another textbook example of a mismatch between assets and liabilities? We will just have to wait for the future to reveal the answer.

Venture capital investors have unfortunately permitted the AI industry to severely distort how it reports its earnings.

To understand the root of this issue, we can look back at the 2010s when SaaS companies established the Annual Run Rate as a popular financial metric. This figure was highly reliable for projecting long term income because it was based entirely on contractual agreements. Forecasting those earnings into the future was completely appropriate, as the revenue was guaranteed for the entire lifespan of the contract.

However, when evaluating AI businesses today, VCs have adapted this concept into what they call Annualized Run Rate. Under this new interpretation, they take short term transactional or API revenue generated over just the previous month, or sometimes even a single week, and project that brief snapshot forward.

It is quite astonishing that investors have allowed this shift in reporting standards to occur, but it has indeed become the reality.

We are truly honored to secure the #2 position in the country.

Ranked #3 nationwide.

Currently, Western society is grappling with a trio of major ideological challenges, namely radicalism, socialism, and doomerism. It is highly common for individuals to be influenced by all three of these mindsets simultaneously. Looking ahead, the ultimate trajectory of the Western world will depend directly on the degree to which these specific movements manage to sway the younger generation.

Roughly ten years ago, I made the strategic choice to cease managing outside money, choosing instead to dedicate my attention to growing internal capital. However, due to the inherently prolonged nature of Venture Capital, bringing VC funds to full realization is a process that spans many years. I am now nearing the very end of this long process, which is why I am distributing my most recent performance summary for Q2. Within the next year or so, I expect to completely wind down these remaining vehicles and officially close this professional chapter.

As I look back, I want to share a few essential lessons I have gathered for both GPs and LPs.

First, your financial reporting needs to be completely straightforward and apparent.

Second, Net DPI is the single metric of true importance. When you optimize for actual returns, you build operational discipline, and that exact discipline acts as your lifeline during difficult economic periods.

Third, it is imperative to benchmark your progress against the Cambridge Index. Measuring yourself against this standard provides the essential motivation required if you truly want to become the GOAT in your field.

At this moment, the AI/SI industry is primarily driven by a CPM approach, where the standard offering is simply the Cost per thousand tokens plus a markup. Looking ahead, the environment will inevitably shift toward a CPA framework. In that future landscape, the focus will turn to task accomplishment, allowing providers to share directly in the generated ROI.

Navigating this shift from CPM to CPA will expose much of the prevalent gamesmanship currently surrounding run-rates, revenues, and gross margins. Stripping away these deceptive practices will inevitably result in a considerable amount of value destruction. We only need to look back at the transitions seen in Web1 and Mobile for clear examples of this phenomenon.

Wishing the best of luck to all the players involved in this evolving space.

Attention Directors, CIOs, CFOs, and CEOs.

Are you familiar with the details of the White House Accord on Super Intelligence? More importantly, what is your strategy for securely integrating Super Intelligence into your corporate environment?

To address these critical questions, 8090 and EY have partnered to develop a novel system named KYSI, or Know Your Super Intelligence. This platform is purposefully constructed to establish a verifiable log of decisions driven by artificial intelligence. It achieves this by recording the relevant evidence and contextual background, while also emphasizing system controls throughout your corporate infrastructure. Ultimately, KYSI empowers companies to maintain rapid technological advancement while simultaneously strengthening oversight.

This new solution supports your business by delivering complete end-to-end traceability, clear policy-to-risk mapping, and comprehensive evidence packages prepared for auditing.

Through this joint effort, we are guiding companies past the phase of simple artificial intelligence trials. We enable you to establish reliable, governable AI frameworks that can confidently scale across the entire organization.

If you would like to explore this further, please get in touch with your dedicated EY Partner or contact us directly at http://8090.ai.

Selecting between an open or closed system is never just a simple black and white decision. Because there is a wide spectrum of access levels available, your final choice should always be driven by the actual job that needs to be done. For instance, relying on a public API makes perfect sense when you need rapid results for low stakes tasks. Alternatively, utilizing a private cloud closed model is the best route when you want to balance high quality with strict control. Finally, if your data absolutely cannot leave your environment under any circumstances, a self hosted open model is the required solution.

The group I work with at @SocialCapital recently outlined this exact framework in our newest Deep Dive on Open vs Closed Models. The insights shared below are taken directly from that presentation deck.

I would like to share my personal perspective regarding artificial intelligence. Please keep in mind that your own viewpoints and experiences on this subject may differ from mine. You can find the video discussion at the following link:

https://youtu.be/iI9uVDF0R4s?is=u4SQViHnMhU6AY5e

Dario recently shared his perspective that the industry needs to decelerate the advancement of artificial intelligence capabilities. By presenting this stance, he is effectively pushing to shut down open source development in order to centralize immense economic and technological influence directly within Anthropic. If you would like to read his complete thoughts on the subject, you can find the original post at https://darioamodei.com/post/we-must-pace-the-frontier

Greetings everyone. I am excited to share that I am establishing a new business enterprise. You can expect this venture to be incredibly significant and to generate immense value, even though it will simultaneously be quite unpleasant and leave a trail of death and destruction in its path. Furthermore, while our upcoming initiatives are designed to be completely transformational, they will also inevitably dismantle society in its current form.

Taking all of these factors into consideration, my primary request is that you purchase the products our organization offers. You can absolutely depend on our team. I sincerely appreciate your time and support.

This certainly gives us something to think about.

This year, CloudBees gathered insights from 200+ enterprise technology leaders, revealing a fascinating contradiction. While 92% of these professionals felt certain their AI-generated code was completely ready for production prior to deployment, 81% of this exact same group experienced an increase in production problems following the release.

Trusting artificial intelligence outputs has officially evolved into a quantifiable business risk. Company boards can now directly weigh the 92% confidence in AI generated code against the 81% production issue rate to gain a clear, realistic picture of their operations. This dynamic raises a vital question for organizations: are development teams merely engaging in unchecked tokenmaxxing, or is there a structured strategy in place that actually delivers results?

As the study points out, and we completely share this perspective, the primary operational bottleneck has shifted away from the initial writing of code and directly toward its governance.

To help teams proactively manage this shift, 8090 Software Factory specifically integrates tests, peer reviews, and formal approvals right into its core features. This system ensures quality control while generating a transparent, easy-to-follow historical record for all stakeholders.

You can find out more details at this link: http://bit.ly/3UWcTQE

It is generally a wise approach to do exactly the reverse anytime someone advises you to simply trust the experts. Think back to the Covid era, when placing absolute faith in specialists resulted in a massive societal deception that devastated our economy and significantly hindered the development of our children. As we move into the age of AI, applying that same blind trust will likely lead to consequences that make our past experiences look entirely minor by comparison. Therefore, I highly recommend that everyone explore these new technological advancements with a sense of amazement, while being sure to maintain a strong foundation of critical doubt.

Understanding what a model means within the banking sector is quite straightforward. It represents the mathematical foundation driving nearly all operational choices, such as determining loan eligibility, calculating necessary capital reserves, and identifying potentially fraudulent transactions.

A significant regulatory shift occurred in April when the Fed, OCC, and FDIC officially retired the framework that financial institutions have relied on since 2011 to validate these mathematical processes. While the updated rules address conventional models, they intentionally exclude agentic and generative AI. The reasoning is that artificial intelligence is advancing far too quickly to be constrained by static guidelines. Consequently, financial institutions are now tasked with independently regulating the most rapidly developing segment of their technological infrastructure. Despite this operational autonomy, banks remain fully obligated to justify their AI governance to examiners whenever requested.

Should such an examination take place tomorrow, the majority of financial institutions would face significant challenges in producing a precise, reproducible log detailing exactly how artificial intelligence was utilized. Tracking the specific methods, the associated artifacts, and every individual decision influenced by AI remains a complex hurdle.

This specific compliance gap is exactly what 8090 Software Factory is designed to resolve. The platform embeds comprehensive evidence and audit trails directly within the software architecture itself. By doing so, the system ensures that every single system modification and AI-driven choice can be reliably traced straight back to a requirement that was explicitly approved by personnel.

Discover additional details by visiting this link: http://bit.ly/46q2SOe

Research conducted by McKinsey and Oxford represents the most extensive analysis ever performed on major IT initiatives, and the findings are quite revealing. After reviewing more than 5,400 projects with budgets exceeding $15M, the data showed that these endeavors typically fall short of their planned value by 56% while simultaneously exceeding their budgets by 45%. Furthermore, the delays in scheduling are just as severe as the financial excesses, frequently pushing the realization of core objectives back by multiple years.

Historically, organizations have tried to navigate these massive IT transformations by making a universally accepted choice and partnering with a firm that no one could fault. Going with a safe option was perfectly logical during an era when simply completing a build was a major question mark. Today, however, AI has completely altered the landscape.

Overcoming these widespread industry challenges requires a system that enables you to clearly document your intentions right at the beginning. By establishing this clarity upfront, you can then leverage AI throughout the execution and verification stages to ensure your project is completed precisely as envisioned. This guarantees that your final deliverables remain on schedule, within scope, and perfectly aligned with your budget.

At 8090 Software Factory, we use this exact approach as the foundation for every client engagement. You can explore the mechanics of our process by visiting this link: http://bit.ly/4gP9bkq

If you are looking for a single visual that highlights a major threat to our current systems, this particular graph immediately comes to my mind. It is highly likely that observers will fail to fully appreciate the compounding of quantum bits moving forward. Ultimately, it is impossible to exaggerate the profound ripple effects this rapid advancement will have on quantum insecure cryptographic methods.