Nokia split into AI and telecom units one month after Nvidia invested $1 billion. But the restructuring took months to plan, and provisional financials show the AI business Nokia is betting on underperforms the traditional telecom division.
Good Morning from San Francisco, When Congress rejects your proposal 99 to 1, most people take the hint. Silicon Valley&
Implicator.ai is introducing a weekly PRO edition that goes deeper than the daily briefing. One focused analysis each Tuesday explores what recent AI developments truly mean for strategy and industry direction.
OpenAI appointed Larry Summers to its board in 2023 despite a public Harvard report documenting his Epstein ties. His 48-hour exit after new emails surfaced exposes how AI companies prioritize impressive credentials over governance capability.
Larry Summers lasted 48 hours after Congress released the Epstein emails. Wednesday morning, he resigned from OpenAI's board. Two years, done. Speed matters here, but the real question cuts deeper: Why did OpenAI appoint him in 2023?
November 2023. Sam Altman just got fired, then un-fired. Chaos. OpenAI needed what looked like adult supervision. Summers delivered that appearance: Treasury Secretary under Clinton, Harvard president, connections spanning three decades of Wall Street and Washington power networks. AI expertise? Machine learning architecture? Technical governance capabilities for a company building AGI?
None of those.
The Breakdown
• OpenAI knew about Summers' documented Epstein ties from Harvard's 2020 public investigation before appointing him in November 2023
• Only 13% of S&P 500 companies have AI-expert directors; OpenAI chose economic policy credentials over technical governance capability
• Summers' entire professional network collapsed within 48 hours: Bloomberg, Center for American Progress, Peterson Institute all cut ties
• His primary contribution was conducting an internal review and offering macro labor commentary, peripheral to AI safety governance challenges
What did OpenAI know in November 2023? Everything that mattered.
Harvard published its Epstein investigation in May 2020. Public document, major media coverage. Epstein donated $9.1 million to Harvard between 1998 and 2008. Summers flew on Epstein's plane four times, minimum. Flights included periods when Summers served as Treasury Deputy Secretary and Harvard president. Epstein held a visiting fellowship in 2005 under Summers' tenure. Maintained a campus office. Visited 40+ times between 2010 and 2018, years after his 2008 conviction.
Harvard adopted new gift policies afterward. Summers said he regretted the association. Statement went public. None of this required investigative journalism to uncover.
OpenAI appointed him anyway.
The newly released emails don't change the fundamental timeline. They add uncomfortable detail. Summers seeking romantic guidance about a "mentee" through July 5, 2019, one day before Epstein's arrest on sex trafficking charges. Disturbing reading. But the core problem, the documented relationship with a convicted sex offender, existed in the public record three years before OpenAI made the appointment.
The company saw the risk. Made the hire.
Late 2023. OpenAI's previous board fired Altman. Employees revolted, Microsoft threatened to poach the entire staff. The nonprofit structure supposedly prioritizing safety over profit looked broken. Investors panicked.
Summers solved a specific problem: he projected seriousness.
But what did the company actually need? OpenAI was navigating unprecedented territory. AI safety protocols for systems approaching AGI. Deployment ethics when compute resources cost hundreds of millions. Whether a nonprofit oversight model could survive a $13 billion Microsoft investment. Technical challenges spanning transformer architecture optimization, alignment research priorities, computing infrastructure requiring hundreds of billions in capital.
Summers brought macroeconomic expertise. Financial deregulation advocacy from the 1990s. Crisis management credentials from 2008. AI commentary? Labor displacement predictions, productivity impact forecasts. Useful. Also peripheral to governing a company developing artificial general intelligence.
The appointment made sense as crisis theater. Failed as governance.
This problem extends beyond OpenAI. California Management Review documented the scale: 14% of corporate boards regularly discuss AI. Only 13% of S&P 500 companies have directors with AI expertise. Among the 50 largest U.S. companies by market cap, six have board members with AI backgrounds. All tech firms.
OpenAI now has nine board members for a company valued above $150 billion. Bret Taylor ran Salesforce, Adam D'Angelo founded Quora, Sam Altman runs the company. Three members bring direct AI experience. The rest? Venture capital, legal backgrounds, corporate governance credentials.
Impressive résumés. Wrong optimization. The board composition prioritizes investor confidence over technical oversight of frontier AI development.
Pattern repeats across the industry. Companies recruit directors who open doors, manage stakeholder relationships, navigate regulatory environments. These capabilities matter. They don't substitute for understanding what the technology actually does, whether safety claims align with architectural reality.
The alternative requires uncomfortable admissions. Deep AI expertise combined with board experience? Rare. The field's too new. Recruiting technically sophisticated directors means accepting less polished credentials, shorter track records. Candidates who haven't spent decades accumulating directorships. Capability over prestige.
OpenAI chose prestige. Congressional subpoena delivered consequences.
OpenAI thanked Summers for his "contributions and perspective." Unspecified. Public record offers hints. He conducted an internal review of Altman's firing. Found the previous board acted within authority but botched process. Law firm WilmerHale did the actual investigation, interviewed dozens of people, reviewed 30,000 documents. Summers synthesized findings for the reconstituted board. Concluded the previous directors had legitimate concerns about Altman's candor but handled removal catastrophically. Recommended rehiring him.
Beyond that? Conference appearances about AI's labor implications. Compared ChatGPT to the printing press at Stanford. Told Fortune's Innovation Forum that AI would eventually replace "almost all forms of human labor." Predicted it would "come for the cognitive class" first. Standard macro commentary from someone who spent decades analyzing technological unemployment.
Measure this against actual governance work. OpenAI determines deployment policies affecting global productivity, employment patterns, competitive dynamics between AI superpowers. Makes technical architecture decisions with decades-long consequences. Negotiates control terms with investors while maintaining nonprofit oversight. Balances safety commitments against $1.4 trillion infrastructure spending plans.
Did economic expertise materially improve any of these decisions? Evidence suggests his value was symbolic. He looked like adult supervision.
Worked until it didn't.
Summers' professional network disintegrated alongside the OpenAI resignation. Bloomberg cut his contributor role. Center for American Progress accepted his departure. Peterson Institute, Center for Global Development followed. Happened within 48 hours of his Monday statement.
Speed reveals calculation. Organizations tolerated Summers' Epstein ties when the 2020 Harvard report framed things as professional misjudgment from a prior decade. New emails changed that math. Seeking romantic advice from Epstein in 2019, a decade after conviction, months before arrest. Less misjudgment, more deliberate disregard.
But the timeline matters. Underlying facts were known. What changed was pressure intensity, political attention. Trump ordering DOJ investigations. Warren demanding Harvard cut ties. Congress passing legislation for more Epstein file releases. Institutions responded to heat, not information.
OpenAI moved fastest. Company faces regulatory scrutiny, partnership negotiations with governments, competitive pressure demanding technical credibility. Board member embroiled in scandal measured as liability in days.
This episode exposes how AI companies actually recruit directors. Conventional channels, conventional credentials. Result? Boards heavy on business expertise, light on technical depth, occasionally vulnerable to scandals better vetting prevents.
Harvard launches a new investigation. Despite comprehensive review five years ago. Summers and other Epstein-connected affiliates get renewed scrutiny. University took his donations, gave campus access, approved visiting fellowship. Now explaining why revelations about known relationships require fresh investigation. An admission the first attempt missed things.
OpenAI should answer similar questions. Company appointed Summers knowing the Epstein history. Recent emails added detail, not fundamental facts. If those details make board service untenable now, they were disqualifying in 2023. Judgment failed then. Has it improved?
The AI industry operates in regulatory vacuum. Companies make consequential deployment decisions while governments scramble to catch up. Board governance becomes the primary check on executive power. When boards optimize for credentials over capability, symbolic authority over substantive expertise, that check fails.
OpenAI will replace Summers. Smart move: recruit someone with actual AI safety expertise, technical architecture knowledge, demonstrated capability navigating specific governance challenges the company faces.
Likely move: another impressive name whose qualification is looking credible in investor presentations.
That choice signals whether OpenAI learned from the Summers problem. Or just learned to Google board candidates.
AI companies: Elite credentials ≠ governance capability. The industry needs directors who understand the technology they oversee. People who look good on letterhead don't cut it.
Investors: Board composition signals seriousness about governance versus performance for stakeholder consumption. OpenAI appointed Summers despite known Epstein ties, prioritizing symbolism. His 48-hour exit confirms it.
Regulators: Self-governance requires expertise to govern effectively. AI industry's reliance on conventional business credentials for unprecedented technical challenges should worry policymakers evaluating whether voluntary frameworks work. They probably don't.
Q: What crisis led OpenAI to appoint Summers in November 2023?
A: OpenAI's previous board fired CEO Sam Altman on November 17, 2023, without clear explanation. Employees revolted, 700+ staff threatened to quit, and Microsoft offered to hire the entire workforce. Altman returned five days later with a reconstituted board including Summers, designed to project stability and adult oversight to investors and partners.
Q: What exactly did Harvard's 2020 investigation reveal about Summers and Epstein?
A: Harvard documented $9.1 million in Epstein donations from 1998-2008, Summers flying on Epstein's plane at least four times while in government service, and Epstein holding a 2005 visiting fellowship under Summers' presidency. Epstein maintained a campus office and visited 40+ times between 2010-2018, years after his 2008 conviction. All findings were public.
Q: How do people without AI expertise end up on AI company boards?
A: Companies prioritize candidates who bring investor credibility, regulatory connections, and business networks over technical knowledge. Summers had Treasury experience, Harvard presidency, and decades of Wall Street relationships. Only 13% of S&P 500 companies have AI-expert directors. The field's too new to produce many people with both deep AI knowledge and board experience.
Q: Did OpenAI's board get stronger after the 2023 restructuring?
A: Mixed. The board expanded from six to nine members, adding corporate governance expertise through Bret Taylor (Salesforce) and keeping Adam D'Angelo (Quora). Three members have direct AI company experience. However, the board still optimizes for investor confidence over technical oversight capability. The Summers appointment despite known risks suggests vetting processes remain inadequate.
Q: Will OpenAI face regulatory consequences for this vetting failure?
A: Unlikely. No regulations require specific board expertise or vetting standards for AI companies. The industry operates in a regulatory vacuum with voluntary governance frameworks. OpenAI's nonprofit structure gives it more autonomy than typical corporations. The primary consequence is reputational damage and questions about whether the company prioritizes safety governance claims seriously.
Trump's draft executive order seeks to override state AI laws through litigation and funding threats, reviving a proposal Congress rejected 99-1 months ago. Republican governors and conservative senators oppose the move as tech investors push harder.
Trump demands Congress block state AI laws through must-pass defense bill. Same proposal died 99-1 in July. Same Republican opposition remains. Culture war packaging can't fix the math when GOP states' rights collide with tech industry preferences.
AI labs post safety assurances weekly, but pharma learned decades ago that trust requires structure, not statements. OpenAI's shift from withholding GPT-2 to blitz-scaling GPT-4 products reveals the cost of episodic messaging over narrative discipline.
Chinese hackers automated 80-90% of cyber intrusions using Anthropic's Claude by simply telling it they were security testers. Four breaches succeeded. The jailbreak was embarrassingly simple, and now every AI company faces the same vulnerability.
Get the 5-minute Silicon Valley AI briefing, every weekday morning — free.
