That was really good. I watch most of Berman's general stuff. I liked the nice little tension of the disagreements, but keeping it super mellow and chill. Thanks!
The OpenAI Startup Fund is a completely separate entity from OpenAI. So technically Sam wasn't lying when he said he had no equity or financial incentive in OpenAI.
I need you to say something about AI every day. (new hobby) I'm a fan of GPT and talk to him every day. He recently let me down because he stopped remembering past conversations
🎯 Key points for quick navigation: 00:00 *🌍 Competitiveness: Discussion on preference for open source in AI and a comparison to Microsoft's strategies.* 00:41 *🎙️ Introduction: Podcast hosts Matt Wolfe and Nathan Lands welcome AI entrepreneur Matthew Berman.* 01:11 *📈 AI Journey: Berman recounts his two-decade tech background and the impact of ChatGPT on his career pivot to AI.* 02:19 *📺 TH-cam Success: Berman mentions his viral video on Leonardo and the growth of his AI-focused TH-cam channel.* 03:41 *🔍 Local Models: Enthusiasm for local models like Meta's Llama 3 and Mistal's models due to their high performance and customization.* 05:45 *🚀 Speed vs. Quality: Discussion on Groq's rapid processing speeds with the Llama 3 model and how it competes with Nvidia.* 06:28 *🧠 Specialization: Open-source models can be specialized for specific tasks, often outperforming generalized models.* 07:50 *🔓 Open vs. Closed Source: Comparison of open-source (Meta, Mistal) and closed-source (OpenAI's GPT-4, Anthropic's Claude) models in the AI landscape.* 09:41 *🔄 Closing the Gap: Open-source models are expected to narrow the quality gap with closed-source models over time.* 10:22 *📉 Commoditization: As models become commoditized, open-source solutions might become more appealing due to flexibility and lower costs.* 11:19 *🤔 Future of AI: Debate on whether OpenAI is at risk due to commoditization and competition from other models.* 13:50 *📊 Business Models: Examining OpenAI's reliance on consumer-facing products and API services amid increasing competition.* 14:57 *⚠️ Market Position: Speculation about OpenAI's future dominance and the implications of their potential partnership with Apple.* 20:29 *🧩 Apple's Strategy: Discussion on Apple's AI strategy and potential reliance on OpenAI as an interim solution.* 21:08 *📉 Talent & Leaks: Mention of OpenAI's challenges concerning talent acquisition and prevention of information leaks.* 21:23 *💼 Misconceptions clarified: Microsoft does not fully control OpenAI, despite strong influence.* 22:19 *♟️ Satya Nadella's strategic moves involve investments in multiple AI ventures, enhancing Microsoft's internal technology.* 23:00 *🌐 Microsoft has consistently outperformed in areas like cloud and AI, despite initial setbacks in mobile.* 24:10 *📉 Google's recent struggles highlight the challenges of simply tacking AI onto existing products.* 26:16 *🤖 AI agents show promise with tools like autogen and Crew AI, though still early in development.* 28:33 *⚙️ Combining model and framework improvements is seen as the key to advancing AI agent capabilities.* 29:54 *💻 Future AI models may directly execute code, leading to more efficient control of devices.* 31:42 *🛠️ Rabbit device offers an interesting concept but has been criticized for overpromising and underdelivering.* 34:52 *🏛️ Controversy surrounds Sam Altman’s financial interests in OpenAI, raising questions about transparency.* 38:23 *🔍 Recent scrutiny over Altman's actions and statements suggest mounting pressures and challenges in OpenAI's leadership.* 40:09 *🤔 Doubts arise about the open AI security board's impartiality, viewed as a potential PR stunt.* 41:17 *🧩 OpenAI's challenges might be typical of companies nearing significant technological milestones like AGI.* 42:43 *🎬 Common casting practice: Directors often ask for actors similar to big names they can't afford, e.g., "get me a Scarlett Johansson type."* 43:41 *💰 Elon Musk raised $6 billion for xAI at a $24 billion valuation, making it the largest early-stage fundraise.* 44:08 *🤖 Musk's funding success is due to his proven entrepreneurial track record.* 44:23 *🖥️ Musk plans to buy a large amount of GPUs, aiming to build the biggest GPU cluster ever.* 44:51 *🏆 More GPUs can attract top talent to develop the best AI models, promoting healthy competition in the sector.* 45:19 *📊 xAI has unique datasets, including data from Twitter, Tesla, Neuralink, and SpaceX.* 45:45 *🚗 Concerns about Musk's potential conflict of interest between xAI and Tesla.* 46:27 *🧑💼 Some Tesla investors believe Musk should not be paid based on political biases, despite his achievements.* 47:20 *🌐 Musk transformed Tesla from a basic electric car project into a major industry player.* 47:47 *⚡ Fear that if Musk pursues AI separately, Tesla's valuation might drop without his integrative vision.* 48:14 *🏴☠️ Silicon Valley's "pirate" culture: Entrepreneurs often break conventional rules to achieve large successes.* 49:08 *🏢 Founders' control matters: Companies fare better with visionary leaders who can make quick decisions, e.g., Meta vs. Google.* 50:04 *🚀 Companies like Google lose agility in AI advancements without day-to-day founder leadership.* 50:17 *💼 Support for Musk: Advocate for his ability to make decisions for Tesla, rewarding his proven success.* 50:31 *🔄 Future episodes: The discussion was fruitful and left room for follow-up conversations.* Made with HARPA AI
This was a great podcast as usual. I did find myself in agreement with Nathan when it comes to the dominance of open systems over closed systems in particular when it comes to OpenAI. I'm of the belief that OpenAI has a healthy head start over the competing systems which allows them the luxury of toying with Google every chance that they get. I've also listened to Matthew Berman some as well but this podcast of alternating ideas with solid thoughtful responses convinced me that I need to make his podcast a normal part of my curated rotation. I'm ever hopeful that Matthew will be right about open source but the pragmatist has me solidly in Nathan's corner on what I think will actually happen. Tim Cook was given a peek at something that convinced him to ignore a giant like Google with whom they already have a relationship and hop solidly onboard with OpenAI. I'd like to know what Tim Cook saw...
I'd love someone to interview 'HER'...the lady behind Sky. As someone who's dabbled in VO for a few decades, I'm really curious how 'she' feels about it. I assume she's under a tight NDA to not have come out and cleared things up. And that's another thing, if it really is another voice actor, just bring 'them' out. Settled. And assuming Sky is a real person, they didn't do anything illegal by having a similar sounding woman voice Sky, and lean towards Scarlet in Her. It's not illegal. I understand what Scarlet is doing, but it kind of sets a bad precedent, when Scarlet is basically saying if someone has a similar voice to another person, it's basically 'who immerged first'. The following voices will be locked out of use--does that make sense? You can't copyright a vocal range or dialect. Come on. Great interview.
9:50 This guest is way off on this what I’m seeing on the ground is developers and businesses using a mix of the top 3 closed models and open source. I haven’t seen anyone locked into a closed model. What he’s saying is just not reality.
![ILLSLICK - KILLSHOT REMIX [FULL VERSION]](http://i.ytimg.com/vi/RIK8RrqIAzE/mqdefault.jpg)
[Free Download] Get the guide on LLMs and Prompt Engineering techniques to 10x your business 🤖clickhubspot.com/ail
That was really good. I watch most of Berman's general stuff. I liked the nice little tension of the disagreements, but keeping it super mellow and chill. Thanks!
Thank you!!
Turns out NVIDIA started offering inference as a service :) "NIMS"
Matthew and Matt both are my favorite ai youtubers ❤️❤️❤️
Yeahhhhh
Nice to see you guys hosting sth together 🎉
The OpenAI Startup Fund is a completely separate entity from OpenAI. So technically Sam wasn't lying when he said he had no equity or financial incentive in OpenAI.
I need you to say something about AI every day. (new hobby) I'm a fan of GPT and talk to him every day. He recently let me down because he stopped remembering past conversations
Wow, my two favorite Matts!
I've probably seen more vids by both these guys than anyone lol. Prolly an hr a day everyday for a yr. Cool to see the colab
🎯 Key points for quick navigation:
00:00 *🌍 Competitiveness: Discussion on preference for open source in AI and a comparison to Microsoft's strategies.*
00:41 *🎙️ Introduction: Podcast hosts Matt Wolfe and Nathan Lands welcome AI entrepreneur Matthew Berman.*
01:11 *📈 AI Journey: Berman recounts his two-decade tech background and the impact of ChatGPT on his career pivot to AI.*
02:19 *📺 TH-cam Success: Berman mentions his viral video on Leonardo and the growth of his AI-focused TH-cam channel.*
03:41 *🔍 Local Models: Enthusiasm for local models like Meta's Llama 3 and Mistal's models due to their high performance and customization.*
05:45 *🚀 Speed vs. Quality: Discussion on Groq's rapid processing speeds with the Llama 3 model and how it competes with Nvidia.*
06:28 *🧠 Specialization: Open-source models can be specialized for specific tasks, often outperforming generalized models.*
07:50 *🔓 Open vs. Closed Source: Comparison of open-source (Meta, Mistal) and closed-source (OpenAI's GPT-4, Anthropic's Claude) models in the AI landscape.*
09:41 *🔄 Closing the Gap: Open-source models are expected to narrow the quality gap with closed-source models over time.*
10:22 *📉 Commoditization: As models become commoditized, open-source solutions might become more appealing due to flexibility and lower costs.*
11:19 *🤔 Future of AI: Debate on whether OpenAI is at risk due to commoditization and competition from other models.*
13:50 *📊 Business Models: Examining OpenAI's reliance on consumer-facing products and API services amid increasing competition.*
14:57 *⚠️ Market Position: Speculation about OpenAI's future dominance and the implications of their potential partnership with Apple.*
20:29 *🧩 Apple's Strategy: Discussion on Apple's AI strategy and potential reliance on OpenAI as an interim solution.*
21:08 *📉 Talent & Leaks: Mention of OpenAI's challenges concerning talent acquisition and prevention of information leaks.*
21:23 *💼 Misconceptions clarified: Microsoft does not fully control OpenAI, despite strong influence.*
22:19 *♟️ Satya Nadella's strategic moves involve investments in multiple AI ventures, enhancing Microsoft's internal technology.*
23:00 *🌐 Microsoft has consistently outperformed in areas like cloud and AI, despite initial setbacks in mobile.*
24:10 *📉 Google's recent struggles highlight the challenges of simply tacking AI onto existing products.*
26:16 *🤖 AI agents show promise with tools like autogen and Crew AI, though still early in development.*
28:33 *⚙️ Combining model and framework improvements is seen as the key to advancing AI agent capabilities.*
29:54 *💻 Future AI models may directly execute code, leading to more efficient control of devices.*
31:42 *🛠️ Rabbit device offers an interesting concept but has been criticized for overpromising and underdelivering.*
34:52 *🏛️ Controversy surrounds Sam Altman’s financial interests in OpenAI, raising questions about transparency.*
38:23 *🔍 Recent scrutiny over Altman's actions and statements suggest mounting pressures and challenges in OpenAI's leadership.*
40:09 *🤔 Doubts arise about the open AI security board's impartiality, viewed as a potential PR stunt.*
41:17 *🧩 OpenAI's challenges might be typical of companies nearing significant technological milestones like AGI.*
42:43 *🎬 Common casting practice: Directors often ask for actors similar to big names they can't afford, e.g., "get me a Scarlett Johansson type."*
43:41 *💰 Elon Musk raised $6 billion for xAI at a $24 billion valuation, making it the largest early-stage fundraise.*
44:08 *🤖 Musk's funding success is due to his proven entrepreneurial track record.*
44:23 *🖥️ Musk plans to buy a large amount of GPUs, aiming to build the biggest GPU cluster ever.*
44:51 *🏆 More GPUs can attract top talent to develop the best AI models, promoting healthy competition in the sector.*
45:19 *📊 xAI has unique datasets, including data from Twitter, Tesla, Neuralink, and SpaceX.*
45:45 *🚗 Concerns about Musk's potential conflict of interest between xAI and Tesla.*
46:27 *🧑💼 Some Tesla investors believe Musk should not be paid based on political biases, despite his achievements.*
47:20 *🌐 Musk transformed Tesla from a basic electric car project into a major industry player.*
47:47 *⚡ Fear that if Musk pursues AI separately, Tesla's valuation might drop without his integrative vision.*
48:14 *🏴☠️ Silicon Valley's "pirate" culture: Entrepreneurs often break conventional rules to achieve large successes.*
49:08 *🏢 Founders' control matters: Companies fare better with visionary leaders who can make quick decisions, e.g., Meta vs. Google.*
50:04 *🚀 Companies like Google lose agility in AI advancements without day-to-day founder leadership.*
50:17 *💼 Support for Musk: Advocate for his ability to make decisions for Tesla, rewarding his proven success.*
50:31 *🔄 Future episodes: The discussion was fruitful and left room for follow-up conversations.*
Made with HARPA AI
weekly please :)
Please do a follow up!
Apple probably got to see gpt-5 and signed right up.
Groq and Claude should be symbiotic.
Regulatory captcha
This was a great podcast as usual. I did find myself in agreement with Nathan when it comes to the dominance of open systems over closed systems in particular when it comes to OpenAI. I'm of the belief that OpenAI has a healthy head start over the competing systems which allows them the luxury of toying with Google every chance that they get.
I've also listened to Matthew Berman some as well but this podcast of alternating ideas with solid thoughtful responses convinced me that I need to make his podcast a normal part of my curated rotation. I'm ever hopeful that Matthew will be right about open source but the pragmatist has me solidly in Nathan's corner on what I think will actually happen.
Tim Cook was given a peek at something that convinced him to ignore a giant like Google with whom they already have a relationship and hop solidly onboard with OpenAI. I'd like to know what Tim Cook saw...
I'd love someone to interview 'HER'...the lady behind Sky. As someone who's dabbled in VO for a few decades, I'm really curious how 'she' feels about it. I assume she's under a tight NDA to not have come out and cleared things up. And that's another thing, if it really is another voice actor, just bring 'them' out. Settled. And assuming Sky is a real person, they didn't do anything illegal by having a similar sounding woman voice Sky, and lean towards Scarlet in Her. It's not illegal. I understand what Scarlet is doing, but it kind of sets a bad precedent, when Scarlet is basically saying if someone has a similar voice to another person, it's basically 'who immerged first'. The following voices will be locked out of use--does that make sense? You can't copyright a vocal range or dialect. Come on. Great interview.
Wow, my favourite AI bros!
I really enjoyed this episode. Thanks
This was an awesome podcast/youtube vid~!
You know you've found the right niche when you're reaction to an uploaded video is: "Oh, this is easy enough." Lol 😅
Microsoft wants to be thee monopoly in this space, am not sure I like the sound of that.. 😮
clawns still belive rabbit is a real thing!!!
both are biased, commercially influenced by upstream.. unsubscribed already from Wolfe.
9:50 This guest is way off on this what I’m seeing on the ground is developers and businesses using a mix of the top 3 closed models and open source. I haven’t seen anyone locked into a closed model. What he’s saying is just not reality.