brew ai
Posts
NVIDIA - the tortoise and the hare story

NVIDIA - the tortoise and the hare story

Can AI Truly Achieve General Intelligence?

Arindam Majumder
May 29, 2024

Welcome AI Brewers,

While OpenAI has been digging for gold, NVIDIA is selling shovels. This phrase perfectly captures last week’s stock market frenzy, as NVIDIA shares gained more than 9% during Thursday trading hours. It seems the market is convinced that increasing computational power is the key to achieving AGI.

In today’s edition:

NVIDIA - the tortoise and the hare story
Can AI Truly Achieve General Intelligence?
AI Spaces Lucrative for Investment from the VC community
Chart of the Week
Epic Tweet
Bonus

Highlight of the Week

The tortoise and the hare story

In hindsight, there wasn't much of a competition initially. But as AI technology advanced and models improved, it became clear that NVIDIA was pulling ahead. Once nowhere near Intel, the leading semiconductor company, NVIDIA has now surpassed Intel by a significant margin. Intel’s market cap is around $181 billion, but NVIDIA has skyrocketed to $2.62 trillion, surpassing the entire stock market value of Germany.

source: MacroTrends

This remarkable achievement is due to NVIDIA's outstanding earnings, which exceeded market expectations with a record quarterly revenue of $26 billion. The company reported a profit of $14.881 billion, a staggering 629% increase from the same quarter last year. As a result, NVIDIA now controls up to 95% of the GPU market.

Now, NVIDIA is trading with a market cap above $2.5 trillion for the first time, making it the third largest tech company in the world, just behind Microsoft and Apple. In a classic tortoise and hare scenario, NVIDIA has proven to be the steady and strategic tortoise, overtaking its competitors with consistent and impressive growth.

NVIDIA Earning Report Details

Deal Flow

Elon Musk's AI startup xAI just secured $6B in Series B funding, with backing from Andreessen Horowitz, Sequoia Capital, and more Link
Scale AI raises $1B in Series F, doubling its valuation to $13.8B. New investors include Accel, Cisco, Intel, Amazon, and Meta Link
AI music maker Suno raises $125M to revolutionize music creation, backed by Lightspeed, Nat Friedman, and more Link
DeepL, the AI translation startup, raises $300M, doubling its valuation to $2B. Led by Index Ventures, the funds will boost sales, marketing, and R&D efforts Link

Events and Updates

Google's AI gave absurd and dangerous advice, like smoking during pregnancy and adding glue to pizza sauce, sparking heavy criticism Link
Microsoft demoed Co-pilot's real-time screen analysis, aiding e-commerce purchases and language adaptation. Team Co-pilot, integrated with Microsoft 365, enhances meetings, notes, and task management Link
Meta introduced a new multimodal model called Chameleon, capable of processing text, images, and audio Link
Adobe Firefly's generative remove feature is coming to Lightroom, enabling users to remove image elements directly within the app Link

Founder’s Corner

Can AI Truly Achieve General Intelligence?

In an era where artificial intelligence (AI) promises to revolutionize industries, the question remains: can AI ever truly achieve general intelligence? Jack Clark, co-founder of Anthropic, provides a nuanced perspective.

Anthropic aims to develop a safe and reliable general intelligence, transforming AI from reactive tools into proactive collaborators capable of complex, open-ended research across various domains.

The journey to general intelligence is marked by intense competition and significant financial investment. Training advanced AI models demands substantial computational power, creating a competitive landscape dominated by a few major players. A significant challenge is extending the context window of AI models, enhancing their ability to manage extensive, multi-step tasks.

Integrating reinforcement learning with language models has enhanced AI's conversational abilities. Future breakthroughs are expected to drive new capabilities and applications.

Clark advocates for robust safety measures and regulatory frameworks to ensure responsible AI deployment. He suggests a balanced approach to open-source AI, fostering innovation while preventing misuse.

Jack Clark’s reflections highlight the inspiring yet daunting quest for general intelligence. The future of AI, through the lens of Anthropic, promises profound advancements balanced by the need for careful and responsible development.

Learn more from him

VC Insights

AI Spaces Lucrative for Investment

Vertical Applications: Michael Eisenberg emphasizes the importance of using AI to transform existing legacy industries. His firm, Olive, focuses on vertical integration, investing in AI applications like maritime AI (Winward) and AI for insurance (Lemonade).
Hardware and Semiconductors: Haime Madis highlights this space as one of the strongest, with companies like Brck and Lambda Labs building cloud infrastructure for engineering and research teams.
Infrastructure and Platforms: Investments in enhanced models for enterprise use, such as those by contextual AI, are seen as reliable, secure, and suitable for sector-specific applications.
Vertical Markets: Haime mentions health tech, fintech, future of work (especially with voice recognition), and autonomous vehicles as key areas for AI applications.
Horizontal Applications: Companies like Dapta, which build automation platforms with quick integration for business applications, are gaining traction.

AI Risks

Data Dependency: Michael Eisenberg points out that the race for large language models (LLMs) is essentially a race to access all the world's data, making proprietary data a key competitive advantage.
Competitive Moats: Without proprietary data, the business moat for AI companies is low, increasing the risk of commoditization.
Overfunded Startups: Jason Calacanis predicts a massive washout of overfunded AI startups due to unsustainable valuations and revenue expectations.

Learn more about AI Trends, Future of AI, Sentiment in the VC Community