University of La Verne

Honored to have been asked to join this fine university as an Advisor and work with school administration and faculty with a focus on AI strategy and implementation. This wonderful university is located just outside Los Angeles and has a laudable community focus with extraordinary programs in education, something near and dear to my heart.

Member of the Board of Advisors

Curiouser AI

Curiouser.AI. Is an AI startup dedicated to understanding and solving the world's most important and vexing problems. We have invented Reflective AI, an AI that uses Socratic reasoning to challenge the user and help them think more deeply. Alice asked, "Who in the world am I? Ah that is the great puzzle." She was right. We all need to find our why.

Co-Founder & CEO 

University of California, Berkeley

Teach entrepreneurship, strategy, marketing and AI to undergrads and grads.

Instructor

hubble

Provide advisory and mentorship guidance to people in my areas of focus and expertise including entrepreneurship, AI, leadership, career, and marketing and branding. This is a paid service but in the event you are struggling in this economy reach out to me privately and we can try and make something work 

Click here to learn more 

https://www.hubble.social/stephenklein

Advisor

Be10x

Working with the wonderful people at Be10x and the Adityas have prepared a series of Masterclasses on Leadership, Corporate Strategy, Decision Making and Analysis. Content includes recorded material as well as live sessions

Masterclass Intructor

POETIC

Thanks to the amazing Harvard Business School Bay Area Community Partners I will be joining a team of brilliant and dedicated HBS alumni to help this extraordinary non-profit find strategic ways to have a bigger impact.

Dentons

At Dentons, the largest law firm in the world, I owned and directed innovation initiatives resulting in the scaling of the world's largest, most innovative infrastructure, technology, and strategic consulting capabilities. I am also a key member of marketing, branding, innovation, and communications team to enhance the legal profession through the development and marketing of Nextlaw Enterprise, comprised of a number of wholly owned innovation and revenue projects-- by designing systems for integrated campaigns across partnerships, advertising, content, community, product, and pricing models.

A few key achievements in this tenure include:
 
• Led the successful launch of the largest Nextlaw Referral Network with 700 firms in 195 countries, Nextlaw Public Affairs Network, a cross-functional network with 200 PR/PA firms, and Nextlaw In-House Solutions, assisting in-house legal organizations in succeeding in a rapidly changing environment.
• Led a company-wide innovation infrastructure platform project which enables tens of thousands of the firms professionals to contribute, share and commercialize innovation and technology initiatives being developed around the world.

Chief Marketing Officer (Nextlaw)

LOYAL3

I co-founded and developed this technology company that allows individuals to invest in leading brands via initial public offerings, stock rewards, fractional ownership, and loyalty programs. I brainstormed with Bain & Company and the Harvard Business School to produce the world's first marketing behavioral currency based on actual stock ownership as a mechanism for businesses to deepen customer relationships through the cognitive science of ownership. Also inventor of the concept that a public company can use its stock as a powerful loyalty currency, in essence shifting customer behavior creating unprecedented levels of lifetime value.

A few key achievements in this tenure include:
 
• Led product development efforts to maximize growth, development, and ROI for three revenue-generating products, generating millions of dollars.
• Elevated NPS values by 20% by delivering social stock distribution and IPO solutions for top clients, including DreamWorks, GoPro, AMC Theatres, Kate Spade, Square, Kraft, and Virgin America.
• Directed initiatives to expand the company to over 150 employees, 16 clients, and hundreds of thousands of users through effective leadership and marketing for the entire organization.

Chief Marketing & Brand Officer, Chief Creative Officer

Too Far Digital Media (Cross Point Ventures)

Executive Vice President, Global Chief Marketing Officer

Line 6 (Yamaha)

Global Chief Marketing Officer

“the only metrics that will truly matter to my life are the individuals whom I have been able to help, one by one, to become better people.” ― Clayton Christensen


In Defense Of Geoffrey Hinton

When Geoffrey Hinton left Google to warn about where AI is heading, the industry eye-rolled. “Alarmist. Out of touch.” 

But that response says more about us than about him. Hinton didn’t leave to sell a product; he left to speak freely about risk. 

Why the engineers attack the Godfather of their own field:

Cognitive dissonance. If your identity and paycheck come from building something, it’s painful to hear its inventor say it could destabilize society. We protect identity by dismissing the messenger. (Festinger, 1957.) 
Stanford University Press

Motivated reasoning. When careers, funding, and stock options reward optimism, our brains cherry-pick the “it’ll be fine” data. (Kunda, 1990.) 

Groupthink in big labs. Tight teams + deadlines + PR cycles = conformity pressure; dissent feels disloyal. (Janis, 1972.) 

Narrative defense. The dominant story is “accelerate.” Slowing down threatens valuations, recruiting, and press momentum. So the system deputizes its builders to swat away caution.

Control illusion. Engineering culture believes every problem has a patch. Warnings about social-scale failure sound “non-technical,” so they’re waved off as FUD.

Here’s the uncomfortable reality: Hinton’s view isn’t fringe. In the largest survey of top AI authors (2,778 respondents), many gave non-trivial probabilities to extremely bad outcomes; a significant share put ≥10% on outcomes as bad as human extinction. 

That’s not certainty—it’s risk worth grappling with. 

And in 2023, hundreds of researchers, CEOs, and Turing winners signed a one-sentence statement: “Mitigating the risk of extinction from AI should be a global priority.” 

You can disagree with “extinction risk,” but you can’t claim there’s no serious debate. 

Meanwhile, prominent leaders like Yann LeCun argue that doomerism is overhyped and that today’s systems are far from human-level intelligence. 

That tension is exactly the point: serious experts disagree, publicly. In such moments, responsible industries amplify caution, not ridicule it. 

The ask isn’t to stop building. It’s to stop pretending that incentives, psychology, and power dynamics don’t shape our judgment. 

If the person who helped build the engine taps the brakes, you don’t floor the gas, you check the brakes.
********************************************************************************
The trick with technology is to avoid spreading darkness at the speed of light

I’m the Founder & CEO of Curiouser.AI. I also teach at UC Berkeley. We’re building unique Generative AI products designed to augment not automate. Our beta is named Alice, give her a try visit us or DM me

Stephen Klein

Founder & CEO of Curiouser.AI | Berkeley Instructor | Harvard MBA | LinkedIn Top 1% Voice in AI | Advisor on Hubble Platform

No alternative text description for this image

FOR IMMEDIATE RELEASE
In a Surprise And Bold Move OpenAI Announces GPT-6.0 Just Days After Its Launch of GPT 5.0

San Francisco - August 9, 2025 — In a bold move that redefines “bold moves,” OpenAI today skipped its planned announcement of GPT 5.5 and announced GPT-6.0, less than 72 hours after releasing GPT-5.0. 

Here is what Sam Altman, OpenAI CEO says we can expect:

Prompt anticipation. GPT-6.0 no longer requires input. None at all. It responds to what you were going to think even if you were not smart enough to think of it in the first place.

"We've learned that not all our users are that smart," Altman explained, " so we've created a way that they can appear smarter than they actually are and do it faster then they could pretend how smart they really never were with GPT 5.0."

Temporal compression. All past, present, and future knowledge is delivered at the same time.

"This is what cubism was to the art world," he said. "We can create simultaneity in space and time itself and what this means is we may even be able to announce GPT 7.0 before we announce GPT 6.0."

"For example," he said, "Yesterday we will announce GPT 7.0 in advance of GPT 6.0 and in fact may even begin to prep our favorite Internet influencers to begin to tease GPT 4.5, again."

Auto Pilot for the Enterprise (APE) This feature can simultaneously auto dial a management consultant, cut a check for $15 million, share a 72 PPT slide deck to a Board of Directors, install a useless pilot, lay people off, announce a company stock buy-back, place some annoying bots between the company and its customers, release large quantities of mediocre marketing content guaranteed to bore customers to death, and issue an AI-First press release within moments of feeling any pressure whatsoever.

Perception alignment. No longer hallucinates; instead, resets reality itself to align with the initial hallucination even when the AI insists that the new reality is not aligned with what it initially insisted was correct.

Given the overwhelming success of GPT-5.0 over the past three days, OpenAI also announced it will be raising capital at an $800 billion valuation, $300 billion more than its $500 billion valuation which was $200 billion more than its $300 billion valuation last week. 

About OpenAI

OpenAI is committed to safely and responsibly accelerating toward a future where AI achieves capabilities previously considered impossible, and then releases the upgrade before anyone finishes reading the announcement.
********************************************************************************
The trick with technology is to avoid spreading darkness at the speed of light
I’m the Founder & CEO of Curiouser.AI. We’re building Generative AI products designed to make your life more difficult by encouraging thinking, elevating you and helping you define your own limits. Our beta is named Alice, give her a try visit curiouserdotai

It Looks Like OpenAI's Days Are Numbered: Here's the data

With GPT-5's recent launch generating more hype, it's time to examine the fundamentals that get lost in the hype

The Financial Reality:

The Math Problem:

They spend $2.25 for every $1 earned, which is unsustainable at scale. 
Let me repeat that: They spend $2.25 for every $1 they earn.

It's the oldest joke in business: When you lose money on every product, you can't make it up in volume

They're projected to lose $14B by 2026 with total losses from 2023 to 2028 expected to reach $44B

Lost $5B in 2024 on $3.7B revenue¹
Projected cumulative losses of $44B through 2028²
ChatGPT fees rising from $20 to $44/month over five years³
Every single user, free or paid, loses the company money⁴

Won't even possibly achieve profitability until 2029⁵ 

The GPT-5 Reality Check:

Despite the fanfare, GPT-5 benchmarks show it's "roughly on par" with competitors⁶

Claims 45% fewer factual errors than GPT-4o⁷, but that still leaves significant hallucination rates and is that claim valid?

The Talent Exodus Crisis:

Only 3 of 11 original cofounders remain⁸
(It's why they left that's disturbing, not that they left)

Key departures include CTO Mira Murati, Chief Scientist Ilya Sutskever, and researcher John Schulman—many joining competitor Anthropic⁹
President Greg Brockman on extended sabbatical through 2024¹⁰

The Competitive Reality:

Meta's Llama 3.1 runs at roughly half the cost of GPT-4 while reaching 1.1B users¹²

Open source models achieving comparable performance

Google, Anthropic, and Meta expanding rapidly while OpenAI deals with internal chaos¹³

The Valuation Question:

$300B valuation despite losing money on every product¹⁴
(Now it claims $500B?)

Requires maintaining 93% annual growth for five years, achieved by only a handful of companies in history¹⁵
Needs $100B revenue by 2029 to justify current metrics¹⁶

The AI revolution is real. But are we funding innovation, or subsidizing a spectacular implosion?
********************************************************************************
At Curiouser.AI, we're building Alice, an AI designed to help you think critically, not just work faster. Because the future needs humans who can navigate complexity, not just automate it.
********************************************************************************
Sources:
¹ LessWrong analysis, March 2025
² Yahoo Finance, October 2024
³ Fortune, September 2024
⁴ LessWrong financial analysis, March 2025
⁵ TapTwice Digital, May 2025
⁶ TechCrunch, August 2025
⁷ OpenAI official announcement, August 2025
⁸ Fortune, August 2024
⁹ Fortune, September 2024; AI News, August 2024
¹⁰ Artificial Intelligence News, August 2024
¹¹ Fortune, September 2024
¹² Foundation Capital analysis, January 2025
¹³ Foundation Capital analysis, January 2025
¹⁴ TapTwice Digital, May 2025
¹⁵ Foundation Capital analysis, January 2025
¹⁶ TapTwice Digital, May 2025

Taplio

Stephen Klein's Linkedin Analytics

What is Stephen talking about?

's Best Posts (last 30 days)

Want to drive more opportunities from LinkedIn?

Taplio

.css-1jgreyo{background-image:linear-gradient(284deg, rgba(39,43,72,1) 0%, rgba(33,47,183,1) 100%);color:transparent;-webkit-background-clip:text;background-clip:text;}Stephen Klein's Linkedin Analytics

What is Stephen talking about?

's Best Posts .css-qqfgvy{font-size:var(--chakra-fontSizes-md);}(last 30 days)

Want to drive more opportunities from LinkedIn?

Stephen Klein's Linkedin Analytics

's Best Posts (last 30 days)