Hard Fork Podcast Summary: Meta Bets on Scale + Apple’s A.I. Struggles + Listeners on Job Automation
Released: June 13, 2025
Hosts: Kevin Roose and Casey Newton
Produced by The New York Times
Introduction
In this episode of Hard Fork, hosts Kevin Roose and Casey Newton delve into the tumultuous landscape of artificial intelligence (AI) within major tech giants and its repercussions on the job market. The discussion centers around Meta's significant investment in Scale AI, Apple's faltering AI initiatives, and firsthand accounts from listeners experiencing AI-driven job transformations.
Meta's Bold AI Investment and Strategic Shift
Meta's Investment in Scale AI At the heart of this episode is Meta's recent decision to acquire a 49% stake in Scale AI for approximately $14-15 billion (04:33). This move signifies Meta's aggressive approach to reclaiming its position in the competitive AI arena. Co-founder and CEO of Scale AI, Alexander Wang, is poised to lead a new AI team at Meta with the ambitious goal of developing superintelligent systems.
Historical Context and Decline in AI Leadership Kevin Roose provides a historical overview, highlighting that Meta was once a leading force in AI research under Yann LeCun's leadership. However, post-2017, following the transformative publication of Google's transformer models, Meta failed to pivot effectively. Internal turmoil and shifting priorities towards ventures like the Metaverse and cryptocurrency led to a stagnation in their AI advancements (07:51).
Casey Newton adds, “Yann Lecun...did not believe in large language models and still doesn't to this day. He is one of the foremost critics and skeptics of this scaling era of large language models” (10:40). This ideological clash resulted in top AI talent departing Meta, further hampering their ability to innovate in AI.
Current Strategy and Skepticism Meta's latest acquisition and recruitment push aims to reverse the declining trend. However, both hosts express skepticism regarding Meta's ability to execute this turnaround. Kevin states, “I am skeptical that this plan of Meta's is going to work” (25:16), citing the challenge of attracting the elite AI researchers who are already well-compensated and committed elsewhere.
Casey concurs, noting the improbability of Meta effectively transitioning Scale AI’s capabilities into superintelligent AI development. She remarks, “There's no clear path from here to the superintelligence Meta is envisioning” (15:12).
Notable Quote:
- Casey Newton at [12:32]: “They started building Llama and made a decision to open source it...it was meant to be a strategic move that would blunt the momentum of OpenAI.”
Apple's Struggles with AI Integration at WWDC 2025
Disappointing AI Announcements at WWDC Contrasting Meta's aggressive AI maneuvers, Apple presented a lackluster AI agenda at its annual developer conference (WWDC 2025). The hosts recount the absence of groundbreaking AI features, with a particular focus on the underwhelming update to Siri.
Kevin remarks, “Unlike last year...this year it [Apple] seems stuck in the past” (28:49). The anticipated AI-driven enhancements, such as Siri’s advanced capabilities to integrate with various apps and perform complex tasks, failed to materialize. Instead, Apple introduced “Liquid Glass,” a superficial redesign of the operating system focused more on aesthetics than functionality.
Criticism of Liquid Glass Design Casey criticizes the new “Liquid Glass” design, emphasizing that it prioritizes looks over usability: “This is a design that is about how it looks. It is not about how it works” (37:39). She contrasts this with Steve Jobs’ philosophy that “design is how it works,” highlighting the disconnect between Apple's presentation and user experience.
Internal Challenges and Research Setbacks The episode also touches upon Apple's internal struggles with AI research. A research paper titled “The Illusion of Thinking: Understanding the Strengths and Limitations of Reasoning Models via the Lens of Problem Complexity” attempts to downplay the efficacy of large language models (LLMs), fueling the perception that Apple is not committed to advancing AI (44:26). This stance mirrors Meta’s earlier reluctance and adds to the skepticism regarding Apple’s future in AI.
Notable Quote:
- Craig Federighi at [32:14]: “We want it the right way... we want to make sure that we have it very much in hand before we start talking about dates for obvious reasons.”
Listener Insights: AI's Impact on Employment
Listener Christian Danielson on Corporate Responsibility and Taxation Christian from Hood River, Oregon, raises concerns about tech executives' lack of concrete plans to mitigate AI-driven job displacement. He advocates for taxing AI technologies to compensate displaced workers and slow AI advancements until policies can adapt (52:52).
Listener Sarah's Experience with AI-Induced Layoffs Sarah, a junior software engineer, shares her ordeal of being laid off as her team was replaced by cheaper human labor, rather than AI. Now evaluating her position in an AI-first company where developers are assessed based on their AI-generated code, she expresses deep frustration and uncertainty about her career prospects (56:01).
Listener Joseph Esparaguara on Integrating AI in Mid-Sized Businesses As the CFO of a home remodeling business, Joseph discusses his efforts to integrate AI within his company to enhance operations without displacing current staff. He faces resistance from employees who are reluctant to embrace AI's broader applications, highlighting the challenges of fostering AI adoption from the ground up (58:50).
Listener George Dilsey on Training Employees Amid AI Advances George, heading the support team at a B2B startup, implements a strategy of rotating support staff through various departments to cultivate versatile skill sets. This approach aims to keep his team employable in an AI-dominated market by developing expert generalists (65:39).
Discussion on AI and Job Automation: Kevin and Casey analyze these listener stories, emphasizing the broader implications of AI on the workforce. They lament the top-down imposition of AI tools without considering the human element, which can lead to reduced employee morale and hinder the development of future leaders. The hosts advocate for bottom-up AI integration strategies that empower employees and encourage collaborative innovation.
Notable Quotes:
- Christian Danielson at [52:52]: “Why it is that the government shouldn't really... tax the technology.”
- Sarah at [56:01]: “I feel terrible for the people just now graduating.”
- Joseph Esparaguara at [58:50]: “If my current team doesn't evolve, I'll be forced to hire different people who will.”
- George Dilsey at [65:39]: “Turned those folks into expert generalists... gaining different skills across the company.”
Conclusion
The episode encapsulates the dynamic and often contentious intersection of AI development within leading tech companies and its tangible effects on the job market. Meta's high-stakes investment contrasts sharply with Apple's stagnation in AI, while listener stories reveal the real-world struggles and adaptations workers face amid rapid technological advancements. Kevin Roose and Casey Newton underscore the necessity for thoughtful, human-centric approaches to AI integration, advocating for policies and strategies that prioritize both innovation and workforce stability.
Notable Quotes with Timestamps
- Casey Newton: “You can always tell something's missing when you get isolated results like AI that's only right for one of your systems.” [00:02]
- Kevin Roose: “A lot of money.” [04:58]
- Kevin Roose: “I am skeptical that this plan of Meta's is going to work.” [25:16]
- Casey Newton: “They started building Llama and made a decision to open source it...it was meant to be a strategic move that would blunt the momentum of OpenAI.” [12:32]
- Craig Federighi: “We want it the right way... we want to make sure that we have it very much in hand before we start talking about dates for obvious reasons.” [32:14]
- Christian Danielson: “Why it is that the government shouldn't really... tax the technology.” [52:52]
- Sarah: “I feel terrible for the people just now graduating.” [56:01]
- Joseph Esparaguara: “If my current team doesn't evolve, I'll be forced to hire different people who will.” [58:50]
- George Dilsey: “Turned those folks into expert generalists... gaining different skills across the company.” [65:39]
Final Thoughts
As AI continues to reshape the technological and employment landscapes, Hard Fork emphasizes the critical need for balanced strategies that harness AI's potential while safeguarding and uplifting the workforce. The episode serves as both a cautionary tale and a call to action for tech leaders and policymakers alike.