Key Moments
E165: Vision Pro: use or lose? Meta vs Snap, SaaS recovery, AI investing, rolling real estate crisis
All-In PodcastKey Moments
Apple Vision Pro shows enterprise potential, Meta leads over Snap, SaaS is recovering, AI investing is complex, and a real estate crisis looms.
Key Insights
The Apple Vision Pro, despite being a V1 product, shows significant promise for enterprise applications, particularly in training and workflow enhancement, echoing the early days of the iPad.
Meta has demonstrated superior performance and strategic focus compared to Snap, leveraging AI, cost-cutting, and shareholder-friendly actions, while Snap struggles with increasing operational costs and a lack of governance feedback loop.
Early indicators suggest a recovery in the Software-as-a-Service (SaaS) market, with accelerated growth in net new Annual Recurring Revenue (ARR) for cloud platforms and companies beginning to beat conservative forecasts.
Venture capital approaches to AI investing are divided, with some waiting for incumbents to dominate, others focusing on open-source models, and many placing bets on foundational models, infrastructure, and applications, though the economics of proprietary models are questioned.
The real estate market, particularly the office sector, faces a significant downturn with substantial value write-downs and debt impairment, impacting banks and pension funds, with a 'pretend and extend' strategy being employed to manage defaults.
While AI's utility may grow infinitely, the economic value of foundational models trained on open-source data is expected to approach zero, shifting value to infrastructure providers and those with proprietary data advantages.
The Apple Vision Pro, despite consumer skepticism and high cost, exhibits industry-leading comfort and integration potential, hinting at a future computing platform, although societal concerns about detachment persist.
The inherent advantage of proprietary data, such as YouTube's vast and multi-modal content repository, could create an insurmountable moat for companies like Google in the AI landscape.
Commercial real estate faces a dual crisis of declining demand (especially for offices) and high financing costs, while multi-family housing, although facing financing challenges, is supported by strong rental demand.
APPLE VISION PRO: A PROMISING V1 FOR ENTERPRISE
The discussion begins with an examination of the Apple Vision Pro. While acknowledging its V1 status and high price point, the hosts see significant potential beyond consumer entertainment. The device's ability to seamlessly integrate 3D objects into reality without causing dizziness, unlike previous VR headsets, is highlighted as a major advancement. Specifically, its application in enterprise settings, such as enhancing greenhouse operations with real-time data, task management, and automated data capture, suggests a productivity leap, potentially '10x'. The training capabilities, using spatial video to demonstrate complex tasks, are also noted as revolutionary for workforce development.
META'S RESURGENCE VERSUS SNAP'S STRUGGLES
Meta's recent performance is contrasted with Snap's. Meta has seen its stock price and profits soar after focusing on AI, cutting costs, and its stock buyback programs, indicating strong shareholder value creation. In contrast, Snap's stock has plummeted, burdened by increasing operational expenses and enormous stock-based compensation diluting shareholder value. The governance structure at Snap, with one CEO holding almost all voting power, removes any accountability or feedback loop for common shareholders, making it unattractive for meaningful engagement or investment. Meta, while also having concentrated voting power, is perceived as more responsive to market feedback.
SIGNS OF A SAAS MARKET REBOUND
Evidence suggests the Software-as-a-Service (SaaS) market may be over its recessionary period. Recent Q4 earnings from major cloud providers like AWS, Azure, and Google Cloud show significant increases in net new Annual Recurring Revenue (ARR), indicating a bounce-back. While the recovery is still in its early stages, other bellwethers are yet to report. Companies are starting to beat conservative forecasts, and while expectations have been lowered, the trend indicates a bottoming out and re-acceleration of growth from a new baseline, suggesting a healthier ecosystem.
NAVIGATING AI INVESTING: MODELS, INFRASTRUCTURE, AND DATA
AI investing is creating distinct camps among VCs. Some believe incumbents like Google and Microsoft will dominate, while others see open-source models winning. A dominant view is that foundational AI models, trained on open data, will become commoditized, driving economic value to zero. Investors are therefore shifting focus to infrastructure providers (like token-per-second services) and companies with proprietary data advantages. While many startups are betting on model performance, the true long-term value may lie in specialized hardware, efficient token processing, and unique, vast datasets that closed systems like YouTube or private enterprise data can offer.
OPENAI'S STRATEGY AND THE RACE FOR AI DOMINANCE
OpenAI's strategy appears to be building dominance through a multi-pronged approach: attracting a massive consumer user base with ChatGPT, creating a developer platform for custom GPTs that leverages network effects, and expanding into enterprise services. The ease of creating custom GPTs and the perceived lead in model quality, even if marginal, are strong advantages. However, widespread adoption for production-level applications remains hindered by slow response times from cloud infrastructure. The long-term economic sustainability of OpenAI hinges on differentiating beyond model quality, perhaps through exclusive data access or superior performance that current cloud providers cannot yet match.
THE ROLLING REAL ESTATE CRISIS: OFFICES AND DEBT
The commercial real estate market, particularly the office sector, is facing a severe crisis. With an estimated 40% of the office market's value potentially lost, significant debt is impaired, impacting regional banks that are major holders of commercial real estate loans. The strategy often employed is 'pretend and extend,' where debt terms are modified to delay defaults, but this merely postpones the inevitable write-downs. This situation poses a threat to pension funds and retirement systems, whose investments are heavily tied to these assets, potentially requiring government intervention to stave off widespread financial instability.
MULTI-FAMILY HOUSING FACES FINANCING HEADWINDS
While office real estate suffers from a dual demand and supply problem, the multi-family housing sector primarily faces financing challenges. Despite strong rental demand and low vacancies, rapidly increasing interest rates have made refinancing prohibitively expensive. Developers who took out construction loans at lower rates now struggle to secure long-term financing at current 8-10% rates, leading to negative leverage and potential unprofitability. For many, this means needing to inject more equity or facing significant losses, contributing to a rolling crisis as debt obligations come due over time.
THE IMPORTANCE OF SPEED AND DATA IN AI DEPLOYMENT
Beyond model quality, the speed and responsiveness of AI systems are critical for user adoption and practical application. Early lessons from Google demonstrated that even milliseconds of delay significantly impact user behavior and engagement. For AI to move from sandbox environments to production, fast and affordable infrastructure is essential. This necessitates efficient hardware and token processing. While many platforms offer AI models, consistently delivering the necessary speed and usability for enterprise-level applications remains a significant hurdle. Companies with proprietary data, like YouTube, possess a distinct advantage in developing truly differentiated and high-performing AI solutions.
Mentioned in This Episode
●Products
●Software & Apps
●Companies
●Concepts
●People Referenced
Common Questions
While the Apple Vision Pro is in its early stages, potential use cases include enterprise applications for training, streamlined data capture in fields like agriculture, and an enhanced movie-watching experience due to its 3D capabilities.
Topics
Mentioned in this video
CEO of Snap, criticized for not listening to the market and for the company's poor governance and financial performance.
Mentioned as the CEO of Meta who made significant changes like a 20% headcount cut and shifted focus to AI, improving company performance.
CEO of Apple, mentioned for the company's impact on Meta's advertising business, and Apple's overall approach to new technologies.
US Treasury Secretary, mentioned as a potential figure to bail out creditors or implement government action in the real estate crisis.
Mentioned in the context of government bonds and held-to-maturity portfolios losing value when interest rates spiked, leading to solvency issues.
Company discussed for its focus shift to AI and significant reduction in headcount, leading to record profits and stock price.
The leading AI company discussed for its potential to maintain a lead in foundational models due to its consumer adoption, developer platform, and ongoing innovation.
Mentioned as an incumbent that will likely win in the AI space and for its cloud infrastructure (EC2, S3) that provided the initial cloud substrate.
Mentioned in relation to GPUs being used by companies like Together AI for AI model training and inference.
Discussed as a massive data repository with video, image, audio, and text content, potentially 300 times larger than Common Crawl, giving it a significant AI data advantage.
Mentioned as an incumbent that will likely win in the AI space and for its TPU hardware and vast data repositories like YouTube.
A company building general use robots, mentioned in comparison to AI and human-robot interaction.
A platform for open-source AI models, discussed in the context of AI investment and the potential for open-source models to drive down the value of proprietary ones.
An AI company mentioned as a competitor in the open-source model space, which is expected to commoditize foundational models.
Mentioned as an incumbent that will likely win in the AI space.
A company whose performance is seen as a bellwether for the SAS industry, showing signs of reacceleration.
Mentioned as a leading consumer-facing AI application, with potential for enterprise use through custom GPTs and team workspaces.
Mentioned as a cloud provider where infrastructure costs for AI can be prohibitive.
Mentioned as a cloud provider whose infrastructure costs can make building production AI apps unfeasible due to slowness and expense.
A company founded by Jonathan Ross, focused on AI acceleration and hardware, discussed in the context of investing in AI infrastructure.
Discussed intensely in the context of a commercial real estate crisis, particularly in the office market, with significant debt and equity write-downs.
An open-source web crawling dataset used for training AI models like GPT-3, highlighted as significantly smaller than YouTube's data repository.
More from All-In Podcast
View all 213 summaries
64 min“This is Bibi’s War” - Harvard’s Graham Allison on the Influences and Endgame of the Iran War
48 minExiled Iranian Prince Reza Pahlavi: Transition Plan and the Fight for Iran's Freedom
2 minPentagon Insider Reveals the “Holy Sh*t Moment” That Caused the Anthropic Fallout
2 minAnthropic vs The Pentagon
Found this useful? Build your knowledge library
Get AI-powered summaries of any YouTube video, podcast, or article in seconds. Save them to your personal pods and access them anytime.
Try Summify free