Reid Hoffman

Who they are

Reid Hoffman is a Partner at Greylock — co-founded LinkedIn in 2003, Inflection AI in 2022 with Mustafa Suleyman, and Prentis in 2026 with Mark Pincus, and hosts the Masters of Scale podcast.

Person

Reid graduated from Stanford with a B.S. in Symbolic Systems, then went to Oxford as a Marshall Scholar for an M.A. in Philosophy — an unusual pairing that shows up in how he talks about technology: always through a lens of systems, networks, and human consequence. Early career took him through Apple and Fujitsu as a product manager before he tried to build SocialNet.com in 1997, an early online social network that didn't survive. He joined PayPal as Co-Founder and Executive Vice President, which put him inside the famous PayPal Mafia network at exactly the right moment. In 2003 he co-founded LinkedIn, building it into the world's largest professional network before it was acquired; he joined Greylock in 2009 when it was already an established early-stage firm, and has been a Partner there since. The founded-company list is long and specific: SocialNet.com (early social/dating network, shut down), LinkedIn (professional network, acquired), Inflection AI (human-computer communication, co-founded with Mustafa Suleyman in 2022), Manas AI (drug discovery with Dr. Siddhartha Mukherjee, 2025), and Prentis (AI lab automating office workflows, co-founded with Mark Pincus and Ritankar Das, 2026 — in talks to raise $100M). The through-line is building network-effects businesses and institutions at the intersection of technology and human coordination — then writing and speaking about the principles behind them. He publishes at reidhoffman.org, contributes to Fast Company and Inc., hosts two podcasts (Masters of Scale on entrepreneurship and scaling; Possible on the future of technology and AI), and has authored books including Blitzscaling, The Start-up of You, The Alliance, Impromptu, and Superagency. His public voice in 2025–2026 has centered on AI safety, AI 'superagency,' and what leaders need to know about the AI transition — with a May 2026 interview with Bernard Marr on the AI advantage as a recent marker.

Company

Greylock's biggest recent move is the July 2026 close of Greylock 18, a $1.5 billion fund dedicated exclusively to early-stage AI-native startups — the firm says it could have raised more but deliberately capped the fund to preserve the model of each partner making only one or two new investments per year, targeting roughly 25 portfolio companies from the fund. The portfolio is producing exits: Chronosphere was acquired by Palo Alto Networks for $3.35 billion in November 2025, and Figma listed on the NYSE at a $13.5 billion market cap and Wealthfront on NASDAQ at a $2.05 billion market cap, both in December 2025. In 2026, Greylock has been active in AI cybersecurity — co-leading a $53 million round for Cogent in February, leading a $45 million seed for Cylake, and co-leading a $60 million seed for Oak in July — alongside broader enterprise AI bets like the $54 million round for Axiamatic in March. In June 2026, Glenn Kelman (former Redfin CEO) joined as Executive in Residence, and Neiman Mathew was hired as a new partner focused on early-stage AI.

Market

Greylock competes directly with Andreessen Horowitz, Sequoia Capital, Kleiner Perkins, Benchmark, Bessemer Venture Partners, and Accel for early-stage AI and software deals in Silicon Valley — all chasing the same narrow set of founders building foundation-model infrastructure and application-layer AI. The firm's differentiation is a founder-centric model with deep hands-on support (in-house recruiting, customer development, and marketing teams) and the Greylock Edge program for pre-seed founders, positioning it against larger multi-stage platforms. AI cybersecurity has become a concentrated thesis, with portfolio positions in Cogent, Cylake, Oak, Tenzai, and 7AI — a sector where regulatory dynamics like the EU AI Act and U.S. CMMC are actively shaping the opportunity.

Network

Hoffman's closest co-founder relationships span several decades and domains: Mustafa Suleyman (Inflection AI), Mark Pincus (Prentis, 2026), Dr. Siddhartha Mukherjee (Manas AI, 2025), and the broader Greylock partnership including Asheem Chandna, Jerry Chen, and David Sze. Bernard Marr has become a frequent amplifier of his AI views in 2025–2026, with multiple interviews and posts featuring Hoffman's takes on AI superagency and leadership. Sarah Guo was a Greylock colleague he appeared with at TechCrunch Early Stage SF 2020.

Mustafa Suleyman· Co-founder, Inflection AI; Partner, Greylock
Mark Pincus· Co-founder, Prentis
Siddhartha Mukherjee· Co-founder, Manas AI
Asheem Chandna· Partner, Greylock
Jerry Chen· Partner, Greylock
Sarah Guo· Former Greylock Partner; GP, Conviction
Bernard Marr· Author/speaker; frequent Hoffman collaborator on AI content

How they likely show up

Long tenure at Greylock since 2009 alongside continuous company co-founding (Inflection AI 2022, Manas AI 2025, Prentis 2026) → operates in parallel tracks simultaneously; doesn't treat investing and building as mutually exclusive.
Symbolic Systems (Stanford) + Philosophy (Oxford, Marshall Scholar) → frames technology problems in terms of systems design and human consequences, not just product mechanics; expect philosophical framing in conversation.
Five named books, two active podcasts (Masters of Scale, Possible), regular contributions to Fast Company and Inc., and a personal site at reidhoffman.org → highly output-oriented, thinks in public, and likely processes ideas by writing and speaking them out.
Hybrid role type (VC partner + serial co-founder) → probably finds purely passive advisory relationships unsatisfying; wants to be involved in the building.
Pattern of co-founding with domain experts (Suleyman on AI, Mukherjee on biomedicine, Pincus on consumer software) → recruits to complement gaps rather than replicating his own profile; values deep domain knowledge he doesn't personally hold.
Departing Microsoft board in 2026 after nearly a decade → willing to end long-standing institutional relationships when the strategic context shifts; not sentimental about affiliations.

Conversation tips

→ Anchor on a specific idea from Superagency or the Possible podcast — he engages far more with people who've read his actual arguments than with people who've read about him.
→ Ask about Prentis specifically — it's his most recent company co-founding (2026, in talks to raise $100M) and the thinking there will be live and unscripted.
→ The Oxford Philosophy background is real, not decorative — if you want a deeper conversation, connect whatever you're discussing to a first-principles question about human agency, network effects, or institutional design.
→ He thinks in network terms: whenever possible, frame problems as coordination problems or network dynamics rather than single-actor challenges.
→ Don't position the conversation as an interview or pitch — he's most engaged when it's a genuine exchange of views, ideally one where you bring a specific take he can push back on.

Toolbox

Openers

Open on Prentis — the AI lab he co-founded with Mark Pincus and Ritankar Das in 2026 to automate office workflows is in talks to raise $100M and almost certainly top of mind; asking what problem Prentis is solving that Inflection AI wasn't signals you've tracked the full arc.
Reference the Greylock 18 deliberate cap at $1.5 billion — the firm says it could have raised more but chose not to in order to preserve the one-or-two-investments-per-partner-per-year model; that's an unusual and specific constraint worth asking about.
Bring up the May 2026 Bernard Marr interview on 'the AI advantage' — he's been active on AI superagency as a public theme and will have refined views on what leaders are still getting wrong about the transition.

Discovery questions

Manas AI (drug discovery) and Prentis (office workflow automation) look like very different bets — what's the common thesis connecting them for you?
With Greylock 18 capped at $1.5 billion and roughly 25 companies, and partners each making one or two investments a year — how do you decide which one or two to take in a year when there are clearly more compelling AI-native founders than slots?
Your public writing on AI superagency argues for a particular kind of human-AI collaboration — how has watching portfolio companies actually build with AI agents over the past year changed or sharpened that view?

Avoid

Don't open with broad questions about his 'journey' or generic takes on AI's potential — he's spent years developing specific, named arguments (blitzscaling, superagency, network effects) and will engage far more if you've done the work to meet him at the level of those specifics.

Make it yours