OpenAI Deal Lets Employees Sell Shares at $86B Valuation
OpenAI’s new deal letting employees sell shares at an $86B valuation prompts questions about how secondary sales work, who can buy, and what employees actually receive through instruments like Profit Participation Units. Commenters split between viewing this as a healthy way to reward and retain staff versus a sign of frothy valuations and “greater fool” dynamics, comparing OpenAI’s prospects and moat to rivals such as Google and Nvidia. Broader themes include concerns over wealth concentration, the long‑term viability of current LLM technology, and whether OpenAI’s nonprofit structure and AGI ambitions are compatible with aggressive commercialization.
How Secondary Share Sales Work
- Several comments explain that employees typically sell in a structured “tender offer,” often via platforms like Nasdaq Private Market or Carta, not through public marketplaces.
- Buyers are usually large or existing investors, not random individuals; money is wired directly, separate from payroll.
- Some anecdotes: companies let employees sell a fixed percentage of vested equity at a set price, sometimes using third‑party valuation, with data rooms and legal review.
- Taxes are significant; some note needing cash to exercise options before selling, or using “cashless exercise” loans.
Profit Participation Units (PPUs) and Windfalls
- New hires often get PPUs instead of traditional stock; these are profit‑linked instruments with a reported 10x return cap.
- Example shared: a mid‑level engineer might have a PPU package valued in the mid‑six figures in the prior tender, implying low‑seven‑figure outcomes at current valuation, before taxes.
- Some question whether PPUs are a “scam” since profits could be reinvested, others note this is analogous to common stock without dividends.
- Unclear details: exact participation limits per employee and all lock‑up terms, though a 2‑year lock‑in is mentioned.
Motivations, Governance, and Altman
- Some argue employees backed leadership largely for financial upside and continued aggressive AI development.
- Others emphasize trust in management’s ability to ship products versus more safety‑oriented board members.
- A minority frame this as part of a broader power play, with concern that employee interests and leadership’s long‑term agenda could diverge.
Valuation, Moat, and Competition
- Opinions split on whether OpenAI’s valuation is justified.
- Pro side: strongest productization, huge revenue growth, first‑mover advantage, user data, and integration with a major cloud give a real moat.
- Skeptical side: core techniques are broadly known, competitors (Big Tech and open‑source) are rapidly catching up; models may be a commodity of “who can wield the most compute.”
- Some see secondary liquidity as normal, healthy diversification; others read it as a possible “top signal” or hedge in case LLMs hit a technical or regulatory wall.
Nvidia, Hardware, and “Shovels vs Prospectors”
- Many compare OpenAI’s risk to Nvidia’s position selling the GPUs everyone needs.
- Nvidia is viewed as having a stronger moat (CUDA, supply, tooling), with customers desperate for alternatives but few real substitutes yet.
- Some predict more competition from custom chips (TPUs, MI300X, etc.) in a few years, but see Nvidia as the archetypal “shovel seller in a gold rush.”
Ethics, Social Impact, and Inequality
- Reactions to the employee payday are mixed:
- Some celebrate well‑paid builders of genuinely useful tools.
- Others lament further wealth concentration and question whether AI’s social and ethical risks are being sidelined by financial incentives.
- Broader worries include AGI’s impact on labor, alignment, and whether society is prepared for rights or moral questions around advanced AI.