The Evolution of AI Models with Naveen Rao of MosaicML
ฝัง
- เผยแพร่เมื่อ 18 ม.ค. 2023
- There are many speculations about the future of artificial intelligence (AI), and in this episode, we hear the opinions and predictions of a player in the inner folds of the AI space. Naveen Rao is the CEO and Co-Founder of the machine learning (ML) training platform, MosaicML, and the former CEO and Co-Founder of Nervana Systems. Naveen shares insight into the thesis behind Mosaic and the practical applications of Large Language Models (LLMs), as well as the Generative Pre-trained Transformer-2 (GPT-2) to GPT-3 transition and the challenge of training models with the constraint of data limits. He predicts the evolution of AI models in terms of quantity, size, and function, and the future of computers in general. If you’re curious as to whether it makes sense to build smaller models or own your own model, this episode is for you. Tune in to hear Naveen’s opinions on the impact of AI on the economy, the danger of centralizing resources, what constitutes sentience, and much more.
Key Points From This Episode:
- Naveen Rao’s educational background and interest in synthetic intelligence.
- What led him to start his first AI company, Nervana Systems.
- The thesis behind MosaicML.
- What MosaicML offers customers.
- What Naveen considers to be 2022’s most exciting breakthrough in AI.
- The innovation of ChatGPT.
- The GPT-2 to GPT-3 transition.
- The challenge of training models with the constraint of data limits.
- Naveen explains the concept of synthetic data.
- He predicts the evolution of AI models in terms of quantity, size, and function.
- Why it makes sense to build smaller models and own your own model where possible.
- Data as a moat component.
- Practical applications of LLMs.
- Naveen’s opinion on whether AI will disrupt the economy or increase Gross Domestic Product (GDP).
- The danger of the centralization of resources.
- How MosaicML is making trading more efficient given the limits AI is facing.
- The efficiency improvements MosaicML customers are seeking out.
- Naveen’s prediction for the rate of cost decline.
- The history of the computer password.
- The future of computers.
- The question of what constitutes sentience.
- Naveen recounts the acquisition process of selling Nervana Systems to Intel.
- How the innovator’s dilemma will play out among competitors in the AI space.
- Naveen’s advice for his past self.
More FYI Podcasts: ark-invest.com/podcasts/
To learn more about ARK: ark-invest.com/
For more updates, follow us on:
Twitter: arkinv.st/3corDiY
LinkedIn: arkinv.st/1f7AiVX
Facebook: arkinv.st/3r4EDOU
Instagram: arkinv.st/39td8bO
Disclosure: arkinv.st/39rzF94 - วิทยาศาสตร์และเทคโนโลยี
ARK is aware of hacked third-party TH-cam channels fraudulently posing as ARK &/or Cathie Wood. These accounts are impersonators & not affiliated with ARK Invest or Cathie Wood in any way. ARK Invest will never use TH-cam or other social media to solicit money, including cryptocurrency. Please report any account claiming otherwise! This channel is the only ARK Invest TH-cam channel. For more information, please view our "What Types of Scams Impersonate ARK Invest and Cathie Wood?" section on our FAQ page here: ark-invest.com/faq/
This might be one of the better FYIs to date. Great information and insight.
00:00:00 - 4th Year Innovation Podcast
00:00:57 - In the Elevator With Mosaic ML's Founder
00:03:19 - Mosaic: The Right Tools to Build Intelligence
00:05:54 - AI at 2022: The Big Picture
00:07:23 - GPT-3: The Step-Function Improvement
00:08:41 - Training Large-Scale Algorithms in GPT-3
00:13:46 - Will Large Models Replace Smaller Models in AI?
00:18:38 - Will Data Dominate the AI Economy?
00:22:34 - Will AI Disrupt the Economy?
00:30:46 - Performance vs. cost in neural networks
00:33:16 - AI Cost Decline 60% Annually
00:39:18 - Will AI Become More Conscious?
00:41:21 - The Transition From Startup to Large Company
00:45:23 - Will the Startup Boom Be Driven by Data or Talent?
00:46:57 - How to Start a Business with Mosaic
00:48:56 - Forward looking statements in this podcast
Autogenerated timestamps
I have to wonder if the prominence of major social media platforms is a big contributor to the fact that they think they've exhausted the extent of capturable data on the internet. Is there data that suggests a slowdown of the rate of new information being created on the internet?
Thanks for creating a list of points. Could you please convert them into links to video timestamps?
00:00:00 - 4th Year Innovation Podcast
00:00:57 - In the Elevator With Mosaic ML's Founder
00:03:19 - Mosaic: The Right Tools to Build Intelligence
00:05:54 - AI at 2022: The Big Picture
00:07:23 - GPT-3: The Step-Function Improvement
00:08:41 - Training Large-Scale Algorithms in GPT-3
00:13:46 - Will Large Models Replace Smaller Models in AI?
00:18:38 - Will Data Dominate the AI Economy?
00:22:34 - Will AI Disrupt the Economy?
00:30:46 - Performance vs. cost in neural networks
00:33:16 - AI Cost Decline 60% Annually
00:39:18 - Will AI Become More Conscious?
00:41:21 - The Transition From Startup to Large Company
00:45:23 - Will the Startup Boom Be Driven by Data or Talent?
00:46:57 - How to Start a Business with Mosaic
00:48:56 - Forward looking statements in this podcast
Giggity.
This topic is highly relevant to my interests.
If the models can generate content that ends up online, e.g. some travel blog is written by AI, then the next model scrapes that content but doesn't know it's machine-generated, so you have this positive feedback loop of AI-generated content training the next model ad infinitum. What happens to the quality of the models? No one is going to add markup to their content to label it as machine-generated to prevent this.
Am I the only one that didn't get a sound track?
Interesting, thanks. Could you cover Edge AI and how that will impact tech. ie I see a company "Brainchip" has already commercialised a NN chip tech (Akida) that has super low power consumption and learns on chip. Game changing!
A truly brilliant vid 🙏
sensor data is no joke. people aren't even touching it half the time. streaming stuff has been hard to even pop on a dashboard. online training of ML models..i don't think I've seen it consumer side - maybe some light bayesian stuff for outlier detection in network sniffers or something. Amazon was doing some interesting thing with calibrating their industrial iot stuff. gah. lots of data.
No sound
Me too.
Sound is ok via chromecast to tv. Quite interesting talk.
I have sound.
First like! Woo Hoo!!!
Staying on trend i see. ChatGPT
ARK is dangerous - Destroyed the value of their investors capital whilst collecting their 2% management FEE!
Wrong. They charge 0.75% fee, and ARKK is up 71% since inception. They have created value for long-term investors.
Lol learn how to invest
ARK Invest forever