Subscribe to the 20VC KZfaq channel for more great interviews: www.youtube.com/@20VC?sub_confirmation=1

I watched Sam talk and now my throat hurts

I've watched a lot of interviews with Altman, you've managed to bring very original and interesting questions that got him to give great answers as well.

I know it’s cringy to just say it out loud but if you get it you get it. The game is changing for humanity. You could see Sam is a little fed up because people he thinks should understand it, don’t. It was obvious when he answered that question about when they would be profitable and he said it was the most boring question he gets asked. It’s all about the efficiencies that happen when everyone has a chief of staff at their disposal. just like the president of the United States. If it can be done over the internet, by a team of super smart people it will be available to us. The game is changing. What happens when anyone can have the world’s most effective therapist on a FaceTime call anytime they want. That’s two years away.

So weird discovering Harry Stebbings has a full body.

## Main Takeaways - At any point, there are only 1-3 things that really matter in a company. - These can change, but identifying them and translating them into teams one of the two most important tasks for effective leadership. - The other task is determining the goal you are running towards-as a far out long-term view. - Then, the process of justifying what the few things that matter are is figuring out the fastest accelerants to reach that point. - The cost of very high intelligence will be driven to nearly zero. - There needs to be a clear path on how better underlying intelligence accelerates the product/company, and this must be expressed very clearly, in order to survive. - Ask yourself if hundredfold improvements in models are something you are excited about-this is a really good delineation.

How did you manage to get @sama to wear a 20vc t-shirt?!

Fantastic Interviewing Harry. Fantastic to reveal such insightful depths.

The rate of adoption for enterprise will be faster than expected... for Microsoft Office users. I'm still confused how OpenAI plans to take AI to enterprise without having Satya eat their lunch.

How did you get Sam to come on your show, let alone wear your Jumper? I watch your pod for nearly 2 years now. The amount of big names you manage to convince to come on your pod despite just few hundreds views per episode is incredible. It would be amazing to know your process. I’m sure your pod will grow over time and wish you best of luck on making it happen!!

Wonderful interview. Great work by all, thank you.

Wow I remember watching a year ago 20VC videos that were in the low hundreds in views and recalling why there were so few views. So glad this blew up!

Great interview! Loved how "light" it was without being dull.

I get why someone like Sam wouldn’t understand why the political/ socioeconomic world feels unstable because he probably has a healthy relationship with social media, but….SOCIAL MEDIA. My lord this is so obviously the cigarettes of our time and the blindness floors me.

Should there be some sort of IP rights for a business idea, for example a technical functionality, for small businesses? Meaning that if a behemoth will "steamroll" you, you could get maybe a compensation? I personally think that if you implement something first, you should get some sort of benefit from that. That would encourage invention. On the same note I think that small competitors should be allowed to release better products. But I do think that a behemoth with the cash should compensate the little fish it is going to eat. Complex topic!

What is the company that they talk about at 23:12? The company that will benefit when the models get much better? CC says Clana? Maybe Klarna Bank AB?

Great interview. A lot of great insight. There are a few other questions I wish you had asked though.

how is this channel not subb'ed by at least 1m yet.... good going Harry!

It's fascinating to hear about the different strategies for building AI models! OpenAI's approach, focusing on continuous improvement with scale, showcases the potential for AI to evolve predictably. Exciting times ahead for AI development! 🔍

What's OAI doing to ensure the data they scrape isn't injected with "poisonous" data? How do you trust a source of truth from which your model is learning from and if your model is relearning from these data sources (which are possibly the output of AI model systems themselves), how do you deal with error propagation within the system?

One thing they miss is that hardware changes so often that previous resources can quickly become obsolete, rendering last year's resources useless. This means that new companies with new strategies, training algorithms, and data can change the game at any time. Since it is all about the latest resources and a group of smart people, the business may not be as solid as you think only with your currently approach.

How do i get the 20 VC jumper or sweatshirt

Why is it called OpenAI? What's open here?

29:42 - Sam laughing at the camera knowing full well something broke when he damn nearly was forced into exile by the board

got to know about Samuel Harris Altman is an American entrepreneur and investor best known as the CEO of OpenAI since 2019. Thank you for Chat GTP Sir!

Becoming VC as well Founder are superpower. When VC learning from their exceptional founders to be next great iteration Founder that's superpower.

Was checking everyday!! Finally!!!

One day Sam, I hope you can enjoy a more normal life, with time to read books and pursue other interests. I deeply appreciate all that you do at OpenAI and for the world. Thank you! 😊

Important episode Harry!

How did you get Sam to wear a 20VC sweatshirt?❤‍🔥❤‍🔥 - Awesome pod!

These interviews are killer. Nice work

I miss Ilya in these type of interviews. He's so captivating.

5:35 Is Mr Lightcap taking out a bottle of Absolut Vodka? Wow, that's something and refreshing if it's not only an Ad placement.

Good evening Harry Sam and Brad Super exciting shared conversation. Think I very much require to be part of something similar, to begin development of an idea much like a 'St AI research type system' to integrate into a brand new creative healthcare delivery service. For the betterment of all. Story very short! Truly grateful for your shared experience and wisdom. 💜

Great interview. I learn a lot, thank you.

🎯 Key Takeaways for quick navigation: 00:00 *📈 Sam Altman discusses two prevailing strategies in AI development: one assumes AI models will not improve and the other bets on continuous improvement.* 00:26 *🚂 Altman predicts OpenAI will outperform startups not anticipating continuous AI advancements.* 01:02 *🧠 The inception of OpenAI was influenced by the successes of deep learning and its scalability.* 01:44 *🌟 Despite skepticism from others, the founders' conviction in AI’s potential and continuous progress drove them forward.* 02:11 *📚 The journey involved evolving their approaches and innovations, underscored by a fundamental belief intheir strategic direction.* 03:21 *🤝 Brad Lightcap details his longstanding partnership with Sam Altman, highlighting their complementary skills within OpenAI.* 04:03 *🌐 Early recognition of OpenAI's unique potential due to its progressive improvement and scalability in AI model development.* 04:45 *🔄 Brad's transition to a full-time role at OpenAI emerged from an initial intent to assist part-time.* 06:21 *🛠️ Altman praises Brad's adaptability and his pivotal role in transitioning OpenAI to a fast-growing business model.* 07:20 *🚀 The agility to undertake new challenges and the vision to integrate and scale business functions are key traits of Brad’s leadership.* 08:52 *🎯 Sam's strength lies in identifying and focusing intensely on the few critical aspects that truly matter for OpenAI's strategic progress.* 10:31 *🧭 Discussing the primary current focus for OpenAI, emphasizing the importance of pioneering in AI and sustaining innovation.* 11:52 *🏗️ A significant barrier to OpenAI's progress would be the lack of computing resources, highlighting the need for sustainable scaling.* 14:28 *📅 The balance between few pivotal strategic decisions and numerous operational decisions shaping the company’s trajectory.* 17:57 *🤖 Altman discusses the future potential for AI to provide inexpensive and abundant intelligence, transforming societal capabilities.* 20:11 *🌍 The long-term value in AI will likely hinge on personalized models that integrate seamlessly into individual lives.* 20:51 *💥 The rapid advancement of AI models like OpenAI's threatens the viability of many startups that do not anticipate continual improvements in AI.* 21:05 *🔄 Discussing two strategies in AI development: building with or without the expectation of continuous AI improvement.* 21:46 *🚜 Highlighting how OpenAI's ongoing improvements to its models can inadvertently "steamroll" startups that do not evolve at a similar pace.* 22:15 *🌱 Startups aligned with OpenAI’s trajectory of model improvement are more likely to benefit and thrive.* 22:57 *📊 A company’s enthusiasm for new AI models is agood indicator of its potential to harness AI advancements effectively.* 23:39 *🏥 AI advancements in the medical field could dramatically improve access to care and save lives, illustrating the high stakes of rapid AI development.* 24:35 *🌐 OpenAI favors iterative deployment of AI models to facilitate societal adaptation and engagement, rather than sudden, large-scale changes.* 25:18 *🤖 Deploying AI tools like ChatGPT publicly has been crucial for raising global awareness and understanding of advanced AI capabilities.* 26:51 *🔄 Feedback from early releases of AI models is instrumental in shaping research and development strategies at OpenAI.* 28:30 *🧠 The primary barrier to utilizing AI in scientific research is the current limitations in AI intelligence, emphasizing the need for smarter models.* 30:08 *🤱 ChatGPT's diverse applications range from aiding research scientists to assisting new parents, demonstrating its broad utility.* 31:19 *📈 Discussing the rapid scaling of OpenAI, driven by the versatile and impactful applications of ChatGPT across various sectors.* 32:14 *🎯 The importance of maintaining a mission-driven culture in tech companies to avoid becoming merely a "resume-building" stop.* 34:07 *📊 Enterprises often underestimate the transformative potential of AI, focusing too narrowly on immediate ROI instead of broader, long-term benefits.* 37:36 *🔄 The challenge for large corporations is to adapt to the rapid pace of AI development and integrate new technologies into established workflows.* 41:59 *💬 Effective communication is crucial for CEOs, especially in startup environments, for tasks like explaining company direction and selling products.* 43:22 *🧓 OpenAI tends to hire slightly older employees, balancing experience with fresh perspectives which often come from less expected sources within the team.* 44:19 *🏢 Emphasizes creating a team environment where ideas are judged on their merit regardless of the source's experience level.* 45:42 *🔄 Experience is valuable in certain roles, but in a new industry like AI, lack of preconceived methods can also be advantageous.* 46:22 *🌐 The lack of established playbooks in emerging fields like AI means that experience isn't always beneficial for solving new kinds of problems.* 47:20 *📈 The biggest challenge for OpenAI in the near term is maintaining top-quality research and product innovation.* 47:34 *🔄 Changed views on the pace of AI adoption in enterprises, expecting faster integration than traditionally anticipated.* 48:27 *🧠 Surprised by the consistent scalability of AI models; increasing model size reliably improves performance.* 49:25 *📚 Wishes for more time to read, recognizing personal sacrifices made for work commitments.* 50:31 *💑 Highlights the importance of communication and support in maintaining personal relationships while managing intense work demands.* 51:42 *🔮 Avoids making long-term predictions about OpenAI's future, reflecting on the unpredictable nature of technological progress.* Made with HARPA AI

29:26 interviewer asks how they scaled so big, so fast. Then the answer is about how useful chat GPT is (distraction). Could the answer lie in how much data was scraped without compensation to anyone?

@33:00 wish he'd have one into what "mercenary" means to him. I worked in automation and got called Mercenery by people that did not want their jobs automated (I was the only one hired for automation at the time, so I basically was there to write code to help free up their time, but they saw it as a threat). Just want to understand their perspective better. I am prepared to be replaced by AI and feel I dont do my job correctly if they still need me around for the same tasks. It's like with humans - you can move up once you train someone else well enough to replace your old job that you're wanting to move up from

Brilliant 💯⭕

What mics are those?

Nice. Been waiting for this episode...

Hearing Sam Altman call this a technological revolution tickles my brain. Good to know all the geniuses in the world are in agreement. Who else is pumped for the future 🤟🏽

Was hoping for a conversation a bit more focused on the tech. Would have preferred much nerdier questions.

I think the best bet is to read over the entire project, then begin. So, where is AI? What's it good for? What can it do? What can it not do? Video games? Can it build has it built a video game?

What startups he is talking about if the base model continuously gets better. What is the need for new standalone apps. Its like generalist will keep getting better in everything then why will we need specialists.

34:40 I think you were trying to say they want to cut cost by removing humans from the equation. How are you guys going to stop AI from removing most people's way of making money?

Awesome interview Harry 👌

I listed to Sam Altman to go to sleep everynight

Thank you so much…one day when I get rich I will pay you for this episode 😅

harry we love your shit dude, nice job on the interview, and nice to have brad in there

Thank you

Been waiting for this one

My llm figured out that each time sam sips his water correlates with his annoyance level

Incredible 20VC shirt with Sama!

I see SO much Henry Cavill in Brad's mannerisms and facial expressions. Just me?

There a three glasses filled water placed at the table… One of them drinks straight out of the bottle, the other one nips out of a plastic cup.

brad looks a bit like jimmy fallon

Great interview!

nice interview but dammmm why hasnt there been a leader since steve jobs who had an ounce of his charisma. lol

Harry, don't fiddle too much with the pen, you are making them and us nervous 😀

180 Studios!! 🎉

Harry was steamrolled by Sam when he said thats a dumb question 😂 Since then poor Harry was defensive whenever asking a question if i am asking a terrible question 😅

The startups have bet on the former category because the latter category requires faith in something we don't have yet ... Of course 95% of people are betting on the former - its a startup, a business, not a pipe dream - people need immediate application. Infact, you wouldn't bet on the former at all if the latter was possible. :)

5:33 Hey dude, easy on that vodka, there is still light outside.

Thank you!!

No quotes from Brad in the intro? 😅

Ask yourself if a 100x improvement in models excites you

14 minutes and 14 seconds is the most I can tolerate listening to Sam speak. It will take me a couple of days to finish watching this interview.

Looks like a bright future for ai

Sam is probably right about getting enough compute, but I hope it doesn't rely too much on TSMC. If China invades Taiwan, the foundries will not remain operational. Not for anyone, not even for China if we are being realistic - since TSMC requires international cooperation to run the foundries.

One day its gonna be revealed that Sam Altman is either an AI with vocal fry setting enabled, installed a voice changing module like Elizabeth Holmes

Did anybody try to count every time these two said "like" in a sentence? I am pretty when you sum it up it amounts to at least half of the whole interview.

Anthropic for the win who could trust any of these guys especially Sam

Where is Ilya?!

Damn.."Lightcap" What a cool name!!

I think they should’ve sat the guests across or perpendicular to each other. This set up has them awkwardly trying to make side eye-contact

Why is Sam deliberately trying to speak in a deeper voice 😂

Great talk thank you

decent ai generated video, very convincing, the audio was slightly robotic

“It would seem to me 🤓”

Sam is good-looking! :)

Could you imagine if the king of the world/universe had a vocal fry? OMG that would be a horrifying dystopia.

Brilliant...

Sam Altman is the God of AI. NOW 👑

about the general macro instability part - put Trump back in a few month this will feel very different.

horrendous headline 😂 if i'm their PR team, this isn't a good look...

This is a masters class in narrative building. Im sure that this has nothing to do with the lawsuit.... 😂

Where is Ilya?

I didn't know Jimmy Fallon was into AI

The steamroll comment is going to come back to haunt OpenAI.

Sam looks sad and concerned 😟

The vocal fry is strong with this one.

It's very telling when Sam says asking about revenue generation is the most boring question possible. Your tech may eliminate millions of jobs, if you aren't going to be making money from it, why do it? Take cost of intelligence to $0 for shits and giggles?

I'm not sure intelligence is "some emergent property of matter or something". That's not quite correct in my opinion. It's correct but it sounds like he assumes spacetime to be fundamental and consciousness to be emergent which is upside down in my opinion. I suppose the statement is true in the context he's saying it...

« Open » AI 😅 became CSS (closed source software) with profits steamrolling… Yay for sure it’s inspiring for shar- investors* going crazy about it

Open AI with worldcoin and paypal, nvidia

He’s doing an Elizabeth Holmes with his voice?

my fear for ai is it will all be backwards engineered by money boys "making the world a better place" but we will all end up jacked into endorphin ports, farmed for our originial thoughts

Thank you sam...❤

This was awesome