An Outlook on Generative AI from Intel Capital Portfolio Companies
Last week, Intel Capital pulled off an incredibly insightful event with some of the top minds in Generative AI (GenAI). The event was co-hosted with our friends at Capgemini, and our conversations took place with great views of the San Francisco Bay and the Giants' ballpark.
Our featured speaker, Dr. Matei Zaharia, Co-Founder of Databricks, creator of Apache Spark, and a UC Berkeley Professor, shared his insights on the future of GenAI in a candid fireside chat, moderated by me. We poked fun at the possibility of a Skynet scenario, and he jokingly shared that he is part of UC Berkeley’s ‘Sky Lab’! We explored the return on investment (ROI) of generative AI, the security concerns associated with this rapidly evolving technology, and examined the growing demand for computing resources, which increased participation among the crowd.
Following the fireside chat, Andreas Sjöström moderated a panel featuring a stellar cast: Robert Nishihara, Co-Founder at Anyscale; Marshall Choy, SVP of Product at SambaNova; Spence Green, CEO and Founder of LILT; and Ajit Patankar, Sr. Director and leader of AI/ML at Juniper Networks. Audience members and panelists alike enjoyed conversations about current trends and where the industry is headed. Sharing insights from what we heard below.
GenAI is rapidly transforming various industries by enhancing efficiencies and driving innovation. As we look toward the future, several key trends are emerging that will shape the landscape:
1. Efficiency and ROI
One of the most significant trends is the focus on improving efficiency. Technologies that streamline processes and enhance productivity are seeing quick ROI. This is particularly evident in industries where time and resource management are critical. For instance, automated content generation tools are helping marketing teams produce high-quality content in a fraction of the time and at a lower cost. According to Dr. Zaharia, these are the areas where ROI is most easily measured because they shorten the time to value, market, or task completion.
In contrast, new and innovative ideas and novel use cases for GenAI often face challenges when quantifying ROI. These ideas, while potentially groundbreaking, require more time for market adoption and validation. Companies investing in such innovations need to be prepared for longer gestation periods before realizing tangible benefits. Ajit discussed the evolution of GenAI from 2022 to the present, starting with ChatGPT for various applications, progressing through phases of experimentation, and evolving towards a multi-platform approach. This involves matching use cases and business requirements to ensure applicability and ROI.
2. Security Concerns
Security is a paramount concern in the GenAI landscape. With the ability to access and analyze vast amounts of corporate data, there is an increased risk of sensitive information being exposed to unintended parties. Ensuring robust security measures and compliance with data protection regulations is crucial. This involves implementing advanced encryption techniques, access controls, and continuous monitoring to safeguard data integrity. One specific caution Dr. Zaharia raised is vulnerability to malicious prompts that access unintended data through the growth of GenAI co-pilots. This is exactly what one of our portfolio companies, Zenity, protects against, and it recently published research on vulnerabilities through prompt injection techniques in co-pilots.
3. Domain-Specific Large Language Models (LLMs)
Unlike general-purpose LLMs, domain-specific LLMs are tailored to specific industries or applications, offering more relevant and accurate outputs. For example, in the healthcare sector, domain-specific LLMs can provide more precise diagnostic suggestions by leveraging specialized medical knowledge. This trend reflects the broader movement toward customization and specialization in AI, through which LLMs can deliver higher value and more actionable insights, driving better decision-making and outcomes.
4. Explosion of Compute and Differentiation
The panel discussion highlighted critical factors driving the GenAI industry: the explosion of and differentiation in computing. Dr. Zaharia noted, "The demand for computational power is growing exponentially, and this is pushing the boundaries of what GenAI can achieve." This surge in computational requirements results in innovations across hardware and software, strengthening more complex and sophisticated AI models.
SambaNova’s Marshall Choy also commented on compute: "Differentiation in compute is not just about raw power; it's about optimizing compute resources for specific AI tasks, "noting that companies are increasingly looking for specialized hardware and software solutions that can efficiently handle their unique AI workloads. This trend is fostering a more diverse and competitive market, with various players offering tailored solutions to meet specific needs. Robert Nishihara from Anyscale explains he spends a lot of time focusing on the compute side, predicting that “…there will be an explosion of new compute accelerators for AI, each with optimizations for specific AI workloads.”
These industry thought leaders left us with some parting thoughts:
- Dr. Matei Zaharia, Databricks: "The future of GenAI lies in its ability to integrate seamlessly with existing workflows and enhance productivity without compromising security. As we continue to innovate, the focus will be on creating AI solutions that are not only powerful but also practical and secure."
- Marshall Choy, SambaNova: "We are at the cusp of a new era in AI, where the differentiation in compute will drive unprecedented advancements. By optimizing compute resources for specific tasks, we can unlock the full potential of GenAI and create solutions that are both efficient and effective."
GenAI trends and thought leaders indicate a rapidly evolving industry poised for significant growth and transformation. The focus on efficiency and quick ROI, coupled with the development of domain-specific LLMs and the explosion of compute, is driving the industry forward. However, these advancements come with their own set of challenges, particularly in terms of security and the quantification of ROI for new ideas.
As the industry moves forward, companies need to strike a balance between innovation and practicality, ensuring that their GenAI solutions are not only cutting-edge but also secure and efficient. By doing so, companies can harness the full potential of GenAI and drive meaningful change across industries.
If you would like to know more about the GenAI companies in the Intel Capital portfolio, and those that focus on securing and enabling AI projects, I would be delighted to chat and share or introduce the relevant tech leaders. And look out for an invitation to our upcoming GenAI events in Washington D.C. this October, and NYC in November!
Our GenAI companies that were showcased at this event include: AI21, Anyscale, Articul8, Bria.ai, Fortanix, Landing.ai, LILT, MinIO, SambaNova, TwelveLabs, Zenity, Zyphra.
