In this episode of All-In, the hosts explore recent developments in artificial intelligence, including OpenAI's GPT-5 release and its performance relative to competitors like Grok 4. The discussion covers AI's economic implications, examining the substantial investments required for data centers and the potential returns, with specific examples from companies like Meta showing how AI implementation affects revenue.
The conversation also delves into US-China trade dynamics, particularly regarding high-tech smuggling and tariff impacts. The hosts examine Apple's business strategy over the past decade, discussing the company's $700 billion in share buybacks and its position in the AI landscape. They consider how Apple might have allocated these funds differently, specifically toward AI development and data center infrastructure.
Sign up for Shortform to access the whole episode summary along with additional materials like counterarguments and context.
Industry experts discuss the evolving landscape of AI systems and their implications across various sectors. The release of ChatGPT-5 brought disappointment, with Gavin Baker noting it's the first OpenAI model that didn't clearly surpass its predecessor. Meanwhile, Grok 4 has outperformed GPT-5 in certain benchmarks, including humanity's last exam and RKGI-2.
Ben Shapiro expresses concern about whether AI productivity improvements will justify the rapid investments in the sector. Jason Calacanis points to the substantial future outlays required for data centers, while Gavin Baker discusses promising economic returns, citing Meta's improved revenue from AI-enhanced ad targeting as an example.
The US-China trade relationship continues to evolve, with tariffs generating $30 billion in July alone and $127 billion in 2025. Gavin Baker suggests these measures might provide valuable experimental data, while Ben Shapiro acknowledges their impact while noting potential economic dampening effects.
The discussion turns to high-tech smuggling, particularly the illegal export of Nvidia GPUs to China through Singapore and Malaysia. Phil Deutch emphasizes the importance of stable policies in fostering economic competition, while David Friedberg suggests China's centralized control might hinder its competitive edge.
The panel critically examines Apple's recent business decisions, particularly its $700 billion share buybacks over the past decade. Gavin Baker expresses frustration with Apple's AI products, noting their deteriorating quality, while Ben Shapiro voices disappointment in the lack of exciting new products.
The experts suggest Apple could have better allocated these funds, potentially investing $200 billion in data centers to compete in AI. While Baker remains optimistic about Apple's future AR glasses, Jason Calacanis expresses skepticism about the company's commitment to AR, citing the lack of developer versions and meaningful progress since initial announcements.
1-Page Summary
The conversation regarding the current state and future potential of AI and language models reveals mixed sentiments. Industry experts discuss the evolving nature of AI systems and their impact on various sectors, including technology, business, and economics.
The release of ChatGPT-5 was greeted with some disappointment because its performance did not significantly surpass that of its predecessors. Gavin Baker remarked that this is the first time OpenAI released a model that was not clearly better than the previous one, marking a potential slowdown in advancements. He also noted a technical glitch that occurred shortly after the release of GPT-5, where a system designed to route to the most appropriate model malfunctioned, which could have affected the initial performance metrics.
Grok 4, another language model, was mentioned as surpassing GPT-5 in certain benchmarks. Grok 4 scored higher than GPT-5 on tests like humanity's last exam and RKGI-2, with percentages of 44.4% and 42% respectively. While GPT-5 may lead in specific chatbot functionalities, Grok 4 exhibits superior performance in other significant assessments.
Regarding the broader implications of AI, Ben Shapiro expressed concern about whether the productivity improvements from AI will validate the rapid and large investments in the sector. He drew parallels with the early days of the internet, noting that it took several years before the internet had a substantial impact on productivity. Jason Calacanis expanded on the investment scale for AI, considering the large future outlays required, especially for data centers. He suggested that the business sector could see substantial Total Addressable Market (TAM) growth due to AI.
Gavin Baker discussed the economic returns from AI investments, mentioning that they are commonly measured through return on invested capital, suggesting a promising trend with initial savings mostly in ...
State and Future of AI and Language Models
As the trade relationship between the United States and China remains fraught with tensions, discussions on tariffs, their impacts, and the smuggling of high-tech equipment take center stage.
Tariff revenue has swelled to $30 billion in July, with a total of $127 billion collected so far in 2025. These measures saw the trade deficit shrink to its lowest since June of 2023, indicating immediate fiscal benefits. Gavin Baker expresses that the experimental approach may yield insightful data, as there is dispute over their efficacy.
Despite expectations of significant retaliation, the U.S. has not experienced much, and Baker suggests this might indicate more leverage than anticipated. Ben Shapiro, aligned with President Trump's use of tariffs against China, also notes their impact. The panel reports a 10% tick up in inflation, somewhat connecting it to the tariffs, and discusses the possibility of tariff revenue decreasing should demand drop. Shapiro also highlights the economic dampening effect of the trade war, suggesting the U.S. economy might perform better without it.
Focusing on illicit activities counteracting trade policies, the panelists mention the smuggling of Nvidia GPUs into China at astonishing rates. Two Chinese nationals were arrested for exporting millions worth of these GPUs to China via Singapore and Malaysia, which underscores the challenges enforcement agencies face. This black market activity indicates a deficiency in controlling such high-tech contraband, reflecting the complexity of the geopolitical tech conflict.
David Friedberg assesses China's centralized control over its economy, suggesting that occasional suppression of entrepreneurship could hinder its competitive edge. Phil Deutch underlines the rule of law as a keystone for innovation and touches on the unpredictability of tariffs, emphasizing the importance of stable policies in fostering economic competition.
Unstable Policies Undermining US Competitiveness With China
Calacanis questions the centralization of power in the U.S. government and how executive orders from Presidents, including tariff decisions, could be creating instability in U.S. global economic positioning. Deutch speaks to the negative impact of shifting policies on America's technological ...
US-China Trade, Tariffs, and High-Tech Smuggling
The panelists scrutinize Apple's recent business decisions, particularly the company's massive share buybacks, and debate whether Apple has been neglecting research and development in cutting-edge fields like AI and AR.
The panel expresses frustration with Apple’s significant investment in repurchasing shares, questioning if the tech giant has neglected innovation as a result.
Gavin Baker is especially vocal regarding Apple's performance in artificial intelligence (AI), pointing out that the company’s AI products are not impressive and seem to be deteriorating over time. Ben Shapiro also voices disappointment, noting that there has been nothing particularly exciting from Apple recently that sparked his interest as a consumer.
The discussion highlights that Apple has repurchased $700 billion worth of shares over the past decade, amounting to more than enough to acquire major companies like Disney, Netflix, Tesla, Uber, Robinhood, and BMW. The panelists hint that with better allocation of these funds, such as reducing buybacks by $200 billion to invest in data centers, Apple could have become a serious contender in the AI sector. The speakers infer that Tim Cook's background in supply chain logistics may have influenced the company to focus less on product innovation.
Looking forward, the panelists discuss opportunities Apple could seize to demonstrate its innovative capabilities once again.
The p ...
Apple's Business Strategy and Lack of Innovation
Download the Shortform Chrome extension for your browser