Jan 29, 2025 · 53m · big-technology
Reid Hoffman: Why The AI Investment Will Pay Off
0:00 / 0:00
▶ Watch on YouTube →
gold bands on the timeline = statements, start to end. Hover to read, click to jump. CC turns on captions
In this episode of the Big Technology Podcast, Reid Hoffman joins Alex Kantrowitz to examine the economic realities, technological frontiers, and societal impact of artificial intelligence. Hoffman defends multi-billion-dollar AI investments, outlines his Super Agency framework for human empowerment, and addresses pressing debates surrounding regulation, global competition, and workforce evolution.
How this conversation actually went
Every chapter scored 0–10 on four independent dynamics. Hover any point for the reasoning behind the score. Alex holds 29% of the talking time here. How this is scored →
Alex as informed peer 4.8 Guest teaching 3.7 Guest disagreement 1.8 Alex pushing back 3.5
speaking balance: gold is Alex, purple is the guest (3 minute bins)
Sharpest disagreement ▶ 13:07 Reid dismisses anti-profit media tropes
Reid forcefully pushes back against the premise that commercialization harms product quality, labeling it a flawed media trope and arguing that capital powers societal progress.
Hardest push from Alex ▶ 44:53 Alex challenges Reid on Inflection deal being a masked acquisitionAlex refuses Reid's benign strategic framing and directly confronts him with the assertion that under any other regulatory climate, Inflection's deal would have been a direct acquisition.
Biggest teaching moment ▶ 4:34 Reid explains growth venture risk-discounting vs DCF modelsReid schools Alex on why standard investment banking discounted cash flow models fail in frontier tech, illustrating how venture math prices downside protection.
Alex holds their own ▶ 22:50 Alex details Chinese chip bypasses and Huawei market shareAlex demonstrates deep subject knowledge by citing DeepSeek benchmarks, Huawei Mate 70 hardware advances, and Chinese government smartphone bans to challenge Western policy effectiveness.
the scores for every segment, with the reasoning behind each
| Chapter | Topic | Alex as informed peer | Guest teaching | Guest disagreement | Alex pushing back | Why |
|---|---|---|---|---|---|---|
| The Scale of AI Investment and Long-Term Value Creation | 5 | 4 | 2 | 4 | Alex opens by pressing Reid on whether massive venture capital investments into AI can mathematically pay off given OpenAI's and Anthropic's multi-billion dollar rounds. Reid reframes the scope, arguing that embedding compute into every device will make today's billions look small in retrospect. | |
| Venture Capital Math, Risk Discounting, and Growth Curves | 6 | 5 | 2 | 5 | Alex pushes on VC fund math, noting OpenAI's estimated $5B loss and the standard 7-year 10x return expectation. Reid counters by explaining growth-stage venture dynamics, risk-discounting, and how baseline asset value lowers the required multiple. | |
| VC Investment Strategies and Historic Precedents | 5 | 5 | 2 | 4 | Alex notes that traditional tier-one VCs sat out recent rounds while Thrive led. Reid explains historical precedents like DST's bold bet on Facebook to demonstrate how non-consensus growth checks win in Silicon Valley. | |
| OpenAI's Mission, Openness, and Iterative Deployment | 6 | 4 | 3 | 5 | Alex questions OpenAI abandoning its original non-profit openness premise, citing Elon Musk's 'ClosedAI' criticism. Reid rejects the outsider definition of openness, clarifying that API access and iterative deployment fulfill the core mission. | |
| Capital Requirements and Balancing Commercialization with Mission | 5 | 4 | 3 | 4 | Alex challenges that heavy investor pressure inevitably leads to user growth-hacking and compromised products. Reid rejects the media narrative that commercial capital distorts missions, asserting capital is necessary to subsidize widespread access. | |
| Microsoft and OpenAI: Partnership and Healthy Competition | 6 | 3 | 1 | 3 | Alex probes Microsoft's hiring of Mustafa Suleyman and emerging competition with OpenAI. Reid, sitting on Microsoft's board, frames coopetition between the two as healthy and mutually beneficial. | |
| Foundation Model Competition and Startup Survival | 5 | 4 | 1 | 4 | Alex asks whether all foundational model companies can survive or if some will collapse under high compute costs. Reid concedes that late entrants to frontier models face severe structural disadvantages. | |
| DeepSeek's Rise, Chinese AI Capabilities, and US Export Controls | 6 | 5 | 3 | 5 | Alex highlights DeepSeek and Huawei hardware advances to argue Western export controls on China have failed. Reid pushes back, explaining that export bans serve as speed bumps rather than total stops to slow down Chinese cluster development. | |
| Nvidia Blackwell Yield Delays and Hardware Horizons | 5 | 4 | 1 | 3 | Alex asks about reports of packaging delays and yield issues with Nvidia's Blackwell chips. Reid provides industry perspective, noting that multi-quarter hardware delays are standard and immaterial across a multi-year venture horizon. | |
| Super Agency and Overcoming Historical Technophobia | 4 | 4 | 1 | 3 | Alex cites survey data showing 61% public distrust in AI. Reid draws historical parallels to initial public resistance toward the printing press, industrial looms, and the internet. | |
| Navigating Labor Transitions in the Cognitive Industrial Revolution | 5 | 4 | 2 | 4 | Alex quotes Yuval Noah Harari's skeptical book blurb to highlight the historical pain of technological transitions. Reid acknowledges the friction but argues AI will act as a cognitive tool to assist displaced workers through the change. | |
| How AI Changes Coding and Software Engineering | 5 | 4 | 1 | 3 | Alex asks if learning to code remains valid advice given AI coding capabilities, and Reid explains how software engineering shifts from syntax memorization to high-level conceptualization and AI orchestration. | |
| Critique of Lina Khan and Antitrust Impacts on Startups | 5 | 4 | 3 | 3 | Alex asks for Reid's assessment of FTC Chair Lina Khan. Reid criticizes her antitrust strategy, arguing that restricting big tech startup acquisitions disincentivizes venture funding and perversely consolidates incumbent dominance. | |
| Inflection AI's Strategic Pivot and Microsoft Licensing Deal | 6 | 4 | 2 | 5 | Alex presses Reid on Inflection AI's unusual talent transfer to Microsoft, asking directly if it was an acquisition masked as a licensing deal to bypass regulators. Reid details the startup's pivot from consumer models to enterprise B2B. | |
| Tech Leaders Navigating the Second Trump Administration | 4 | 2 | 2 | 4 | Alex asks about Silicon Valley tech executives courting Donald Trump ahead of his second term. Reid offers a diplomatic answer focused on pragmatic national economic competitiveness, deflecting personal political speculation. | |
| Elon Musk's Political Proximity and OpenAI Dynamics | 5 | 2 | 2 | 3 | Alex questions whether Elon Musk will leverage his Trump administration ties to target OpenAI's for-profit restructuring. Reid downplays the risk as pre-inauguration speculation and cites public conciliatory tweets between Sam and Elon. | |
| Mastering AI Prompting Through Role-Taking and Personas | 3 | 4 | 1 | 1 | Alex sets up a lightning round on voice interfaces and prompting advice. Reid explains how assigning specific professional personas and contrary advocate roles unlocks deep reasoning from LLMs. | |
| Episode Conclusion and Book Recommendation | 1 | 0 | 0 | 0 | Alex wraps up the conversation, plugs Reid's book Super Agency, and closes the episode. |
Statements from this episode (35)
Kantrowitz: OpenAI raised largest VC round in history at $6.6B
“OpenAI raised the largest VC round in history last year, a 6.6 billion.”
Kantrowitz: Anthropic raised $4B and is targeting another $2B
“Anthropic raised four billion and is on track to raise another two billion.”
Hoffman: Current AI investments will seem small in retrospect
“When you think about what AI is going to bring to the transformation of human life and human work, the investments will, I think, seem small. In retrospect”
Hoffman: Everything with a compute unit will become smarter with AI
“When every single thing that has a compute unit in it, and it's not just phones and computers, which will be huge, but you know, speakers and lights and Thermostats and cars and, you know, planes, trains, and automobiles. Everything is going to get smarter.”
Hoffman: Silicon Valley dominates by ignoring traditional DCF analysis
“This is actually one of the reasons why Silicon Valley tends to create so many of the technology companies that dominate the globe, because if you take an overly kind of classic DCF analysis, business school prof analysis, et cetera, which many, much of the wo…”
Hoffman: Greylock's Series A in Airbnb bled money before succeeding
“Greylock investment in Airbnb, which I led the series A of, had that similar characteristics of just bleeding money, and then later being the amazing business we see today.”
Hoffman: Late-stage AI rounds do not require 10x returns to succeed
“The investment that investors made last year is kind of more of a growth venture round relative to, it doesn't necessarily require a 10 X. I mean, you know, we have trillion dollar companies now, so, you know, trillion dollar companies are definitely within fe…”
Hoffman: DST Won Big on Facebook by Offering $10B Post With No Caveats
“In Valley history DST led around in Facebook when the traditional funds were all like, oh, we'll only do a three to five billion pre, you know, with lots of caveats and everything else. And DST came in and said, we'll do a ten billion post and no caveats. And …”
Hoffman: Investing Billions Works When Chance of Losing Capital Is Extremely Small
“Look, you're trying to hit a return multiple, but it's much easier when you know, when you believe with some vigor and evidence that your chance of going below one X is very, very small.”
Hoffman: Rival VCs Called Greylock Crazy for Backing Facebook at $500M Valuation
“Speaking for Greylock, you know, when Greylock invested at Facebook at a five hundred million pre-money valuation, literally all of the companion VC firms were, oh, that's crazy, this college website, Greylock's demonstrating that it's past, past time, et cete…”
Hoffman: OpenAI meant open API access, not open source models
“In terms of being open, there's lots of different parameters of open, and open AI has been, you know, working to have, like, open, you know, kind of you know, with, obviously, identity and registration for safety, but open AI, API access. It wasn't called open…”
Hoffman: Elon Musk approved OpenAI's closed-model strategy as co-CEO
“Elon was part of that decision for you know, as kind of co-CEO in terms of making that decision. And so that's how open AI has always operated.”
Hoffman: OpenAI Offers Many Products Below Cost to Maximize Adoption
“And OpenAI is actually, you know spending a bunch of money making it free or subsidized to get the widest possible exposure, engagement in the, and learning about these products and being able to use them. You know, a lot of OpenAI products are offered at belo…”
Hoffman: History Will View Microsoft-OpenAI as an Epic Partnership
“I think when the history books are written. I think this will be one of the kind of epic, ah, partnerships that, you know, kind of, ah, plays out in the, into the field”
Hoffman: Some foundation model efforts will fail, especially late entrants
“Well, I do think that there will be efforts at foundational models that won't work. And especially I think late entrance like late entrance have a harder, you know, catch up routine and a lot of kind of expense to go to it”
Hoffman: China will produce multiple strong contenders in global AI
“China and multiple companies within it, because, you know, I haven't evaluated yet. Minimax, which also just came out recently, but will have a number of very strong contenders for this in the world technology landscape.”
Hoffman: US export controls are right response to Chinese mercantilism
“And I do think that the, it's the right thing for the U S to be doing, given that China is generally taking a kind of a mercantilist competitive position. And I think that's something that the U S and other you know, kind of countries Europe and everything els…”
Hoffman: Nvidia Blackwell Delays Would Slow AI Development But Not Stop It
“If they can't get their year olds up and produce it, that will be a, you know, concerning thing for NVIDIA and will slow down some of, you know, kind of AI's development because the faster chips in a dense environment is actually, in fact, instrumental and hel…”
Hoffman: 6-12 Month Delays Are Normal for Nvidia and Chipmakers
“It's actually not atypical. For either NVIDIA or chip manufacturers to sometimes be, you know to have six to 12 month error bars in, in the stuff they're doing”
Hoffman: Distrust Is the Default Initial Human Reaction to New Technology
“Very much broadly, most human beings start with a new technology with distrust.”
Hoffman: AI will transform jobs and replace workers who do not use it
“A classic one with AI is say, well, this is going to change a lot of jobs. And it will, cause it's a kind of a new tool set, like a new tool set. Like if you say, Hey, I'm a professional today and I don't use smartphones and I don't use computers. It's like, n…”
Hoffman: Luddites destroying power looms would have doomed their descendants to poverty
“If the Luddites had successfully destroyed all the power looms, And said, well, no, we want to have independent weaving, then that would have meant the, their children, their grandchildren, et cetera, would have been kind of doomed to poverty and less relevanc…”
Hoffman: Rigid coding syntax will become less relevant than conceptualizing software
“You'll need to understand the concepts of kind of code, but AI will be the, Kind of generation a massive generator, a massive quality assurance checker, a bunch of other things. You'll still want to be learning the kind of, as it were, the pattern of the compu…”
Hoffman: Everyone will have an AI coding assistant on PC and phone
“With the AI revolution is all of us will have a coding co-pilot and assistant on our PC, on our phone, that will help us do these things, and so when we think about, like, you know, how do we accomplish some kind of professional task, how we do information gat…”
Hoffman: Precautionary regulation causes Europe to lag the US and China
“The problem with most people's default modes of regulation, including a lot of regulators and including a lot of European regulators is you say, well, our job is to protect bad things from happening. And by the way, the simplest way to do that is to prevent an…”
Hoffman: Iterative deployment with small errors creates a safer technological future
“What you want is you want small errors and small problems that you're learning, hence iterative deployment, that you then fix as it gets to scale, and that's the way you create a safe future.”
Hoffman: FTC Chair Lina Khan's tech M&A crackdown concentrates Big Tech power
“You know, she misunderstands, like, kind of the role of how to stop the aggregation of power in large tech companies, because she says, well, if she stops large tech companies from buying small startups, what that means is venture capitalists, like myself, and…”
Hoffman: Lina Khan Did a Great Job on Drug Pricing and Non-Competes
“Now she's done a great job on like drug pricing and Anti-competes and a bunch of other things.”
Hoffman: Independent startups cannot economically build frontier consumer AI models
“The basic thought from an inflection standpoint is that the original model wouldn't really work, which is building frontier models for doing a consumer agent that the cost curves to the revenue curves and what you would have to do wouldn't work as a startup.”
Hoffman: Inflection deal with Microsoft used non-exclusive licensing and staff transfers
“We could make this deal work You know economically by getting this kind of big licensing deal and ability to kind of do a non-exclusive inclusive license of the technology and bring over a set of the team that really wants to continue to be working on the cons…”
Hoffman: Inflection-Microsoft deal could have been an acquisition under different regulation
“It certainly could be and by the way, inflection, you know, who knows, maybe some year will still be acquired.”
Hoffman: Global tech companies must maintain strong ties with home governments
“One of the things about technology is a more important global business, and if you're doing global, you have to have Good relationships with multiple governments, including, of course, your home country and I think that that's, you know, part of what you're se…”
Hoffman: Musk agreed with Sam Altman not to target OpenAI politically
“Sam has tweeted that you know, he doesn't anticipate Elon, you know, engaging in that kind of behavior, and Elon has retweeted those sections saying, yep, that's totally right.”
Hoffman: Voice will very soon become the main computing interface
“A hundred percent. One of the party tricks I do is I bring out my phone and put ChatGPT on audio mode to show people. Because part of when we learn to use phones or PCs, we've learned, you know, with GUIs and everything else, we've made it softer, but there's …”