AI Leaders to Watch

Also: Apple is promising personalized AI in a private cloud. Here’s how that will work.

Author

AI KATANA
June 11, 2024

Good morning,

In today's newsletter, we spotlight the transformative players in the AI sector for 2024, as identified by Bain & Company, from established giants to innovative disruptors shaping various industries. We delve into the environmental and financial costs of Google's expanding AI capabilities, highlighting the significant energy demands and the strategic moves by tech companies to mitigate these impacts. Additionally, MediaTek's collaboration with Microsoft marks a significant push into AI-integrated laptops, challenging the current market leaders. Apple, meanwhile, unveils its privacy-centric AI services, emphasizing on-device processing to safeguard user data. We also explore Mistral's new AI model for code generation, Codestral, which promises to boost developer productivity across multiple languages. Lastly, we examine the leading venture capital firms fueling the rise of generative AI startups, spotlighting those driving the next wave of technological advancements. Dive in for a comprehensive look at the dynamic forces and emerging trends in AI today.

Sliced:

  • 👀 AI Leaders to Watch

  • 🌱 What Do Google’s AI Answers Cost the Environment?

  • 💻 MediaTek designs Arm-based chip for Microsoft's AI laptops, say sources

  • ☁️ Apple is promising personalized AI in a private cloud. Here’s how that will work.

Bain & Company has highlighted key AI companies to watch in 2024, focusing on firms driving significant advancements in various applications. These companies are categorized into "must-know leaders" and "emerging disruptors" based on factors such as funding, market traction, and innovation. Among the must-know leaders are Glean, providing AI-powered enterprise search; Harvey, tailored for legal services; and Synthesia, known for its lifelike avatars in multimedia applications. Emerging disruptors include Hippocratic AI, focused on healthcare, and Perplexity AI, a consumer search engine. This selection reflects the evolving landscape of AI, emphasizing solutions that cater to specific industry needs and operational challenges. The companies listed have demonstrated substantial growth, securing significant investment and recognition within their respective fields, positioning them as influential players in the AI sector for the coming year.

Google’s ambitious expansion of AI-powered search, set to reach a billion users by the end of 2024, comes with significant environmental and financial implications. Generative AI, which underpins Google's "AI Overviews," requires substantially more energy compared to traditional search methods. Researchers estimate that generating text via AI can consume up to 30 times the energy of simple data retrieval. This heightened energy demand translates into increased greenhouse gas emissions and water usage for cooling data centers. For instance, the daily use of a large AI model can emit as much carbon dioxide as driving a car for 49 miles. Financially, each AI-generated response costs Google approximately ten times more than a standard search, potentially adding up to billions annually. To support these AI services, tech giants like Google and Microsoft are investing heavily in new data centers, which are expected to double in energy usage by 2026. Despite efforts to shift towards renewable energy, challenges remain in aligning data center energy needs with the fluctuating availability of sustainable power sources.

MediaTek is developing an Arm-based chip for Microsoft's AI laptops, poised to challenge Apple's dominance in the space. This move aligns with Microsoft's strategy to optimize Windows for Arm technology, aiming to deliver advanced AI capabilities on personal computers. Scheduled for release after Qualcomm's exclusivity deal with Microsoft expires next year, MediaTek's new chip leverages Arm's pre-designed components, expediting development. This effort underscores Microsoft's broader initiative to integrate AI into consumer computing, potentially shifting the PC market landscape traditionally dominated by Intel and AMD. With both MediaTek and Nvidia collaborating on different projects involving Arm-based designs for Windows, the competition in the AI-driven PC segment is intensifying as companies race to provide robust AI experiences on personal devices.

☁️ Apple is promising personalized AI in a private cloud. Here’s how that will work.

Apple's latest venture into AI, unveiled at their Worldwide Developer Conference, is "Apple Intelligence," a suite of personalized AI services designed to run primarily on the user's device. This approach emphasizes privacy, ensuring that sensitive data remains secure and is only used for the task at hand. When necessary, AI tasks will be processed in a private cloud environment where data is encrypted and not retained post-task. Apple's method contrasts sharply with other tech giants like Google and Meta, who rely heavily on cloud-based data storage and processing. Apple's AI services, powered by their advanced M1 chips and future iterations, aim to handle most computations locally on iPhones and Macs, minimizing data exposure to the cloud. This strategy aligns with Apple's hardware-centric business model, which prioritizes user privacy and security over data-driven advertising revenue. The introduction of Apple Intelligence, starting with the iPhone 15 and new macOS Sequoia, signals Apple's commitment to enhancing user experience with AI while maintaining strict privacy standards, positioning it as a leader in privacy-focused AI innovation.

🛠️ AI tools updates

Mistral AI has introduced Codestral, a new AI model designed for code generation, which aims to enhance developer efficiency across various programming tasks. Supporting over 80 languages, including Python, Java, and C++, Codestral is notable for its large context length of 32k tokens, enabling it to generate more detailed and coherent code solutions. It has outperformed competitors like CodeLlama 70B and DeepSeek Coder 33B on benchmarks such as HumanEval for Python, achieving 81.1% accuracy in code generation. Integrated with popular development tools like VSCode and JetBrains, Codestral simplifies coding by providing robust capabilities in code completion, function generation, and test writing. Although it requires significant computing resources, it’s available under Mistral's Non-Production License for research and testing via HuggingFace. This model exemplifies Mistral’s commitment to advancing AI-driven development tools and positions Codestral as a powerful asset for developers seeking to streamline their coding processes.

💵 Venture Capital updates

In the rapidly evolving landscape of generative AI, venture capital firms have been pivotal in propelling startups to prominence and substantial valuations. Among the most active investors, Sequoia Capital stands out with significant investments in trailblazers like OpenAI and Hugging Face. Andreessen Horowitz is aggressively backing ventures such as Mistral and Pinecone, reinforcing its commitment to dominate the AI sector. Pioneer Fund targets Y Combinator alumni, supporting innovative companies like BoothAI. Soma Capital, with its global reach, has fostered startups like Cognition Labs and Imbue. Khosla Ventures, spearheaded by AI advocate Vinod Khosla, continues its legacy with investments in diverse AI applications, including Replika and SketchPro. Alumni Ventures leverages university connections to fund significant players like Cohere and Lambda. Lightspeed's portfolio boasts Stability AI and Snorkel AI, while Gaingels focuses on inclusive growth with ventures like People.ai and Shield AI. Tiger Global makes substantial investments in leaders such as OpenAI and Cohere, alongside operational tech like OctoAI. Lastly, Greylock Partners supports forward-thinking enterprises, reflecting a broader trend where AI valuations soar, driven by robust investor confidence and a market hungry for innovation.

🫡 Meme of the day

⭐️ Generative AI image of the day