Out with Lawyers, Guns, and Money in with Silicon, Talent, and Data

[The image above is generated by Midjourney, a generative AI platform for creating images. The prompt I used to create the image is listed at the end of this email.]

During the golden age of Silicon Valley, capital reigned supreme, driving the valuation of Web 3 startups to unprecedented heights. However, the landscape has transformed. Artificial Intelligence (AI) is now the new wave, following the slower rise of Web3. Unlike the crypto industry, fueled by enigmatic figures and blockchain prodigies, this new race is led by a different breed of ARMs dealers (yes, that’s a processor pun) —those armed with processors, data, and AI talent. Now GPUs, proprietary domain-specific data, and skilled data scientists are reshaping the competitive technology landscape.

At the heart of this revolution are GPUs (Graphics Processing Units). Unlike the Central Processing Unit (CPU), which focuses on general-purpose computing tasks, GPUs excel at performing parallel calculations on large sets of data, making them ideal for crunching machine learning algorithms. Nvidia, the dominant supplier of silicon for AI applications, stands at the forefront with an impressive 80% control of the GPU market. Nvidia recently achieved a staggering $1 trillion valuation following its latest earnings call, a remarkable testament to AI's growing importance in the tech industry.

In the short term, the frontrunners will be those who build AI platforms to facilitate the deployment and scalability of this transformative technology for early adopters. Unlike traditional venture capitalists who primarily offer capital and business guidance, tech giants bring immense computing power and hardware to the table. Google, for instance, recently led a $1.5 billion investment in Runway, an AI video-editing startup. As part of the deal, Google extended cloud credits and financial support to Runway (Runway was used for video effects in the Oscar-winning movie, Everything Everywhere All at Once). With its long-standing focus on AI infrastructure, Google is a leading provider in this space, having developed dedicated chips and infrastructure tailored for artificial intelligence workloads since 2006.

CoreWeave, a cloud computing company specializing in scaling GPU-intensive workloads, has recently concluded a successful $221 million funding round with Nvidia's participation. CoreWeave's claim of "autoscaling" within three seconds showcases the power of Nvidia's HGX H100 server platform that boasts a 9x increase in processing over the previous generation of GPUs. This partnership demonstrates the symbiotic relationship between GPU suppliers and infrastructure providers, cementing the foundation for future growth in the AI landscape.

The true game-changer, however, lies in AI's ability to level the playing field for companies that may lack technological expertise. As AI technology advances, every programmer will become a 10x programmer, breaking a bottleneck for scaling technology companies. The skills that will be in demand will be their adeptness at debugging and integrating code generated at lightning speed by AI-powered tools such as Github's Copilot, Amazon Codewhisper, autonomous AI agents, and text-to-code platforms such as OpenAI Codex. These tools will streamline development workflow and unlock exponential gains in productivity.

While AI democratizes programming talent, the true advantage lies in companies that can leverage their domain-specific data to train their models. Access to data that can teach these models to excel in specific markets will become a critical competitive differentiator. Bloomberg, the financial information giant, has already embarked on this path by training their BloombergGPT with extensive financial data. The next wave will see companies utilizing their data to train medical and research models and every other field of endeavor, accelerating the development of products and services beyond general-use AI.

Once the dust settles and major architectural deals are finalized, we can anticipate a Cambrian Explosion of startups. These agile companies will employ capital more efficiently and move at an accelerated pace, thanks to the power of AI.

The competitive landscape has fundamentally shifted thanks to AI. Those with access to GPUs, data for training models, and AI talent now outweigh traditional capital considerations. Companies that effectively harness these resources will propel enterprise users forward. It will be a crazy, fast ride for most of us.

Tip of the Week: Using ChatGPT Plugins

When ChatGPT first emerged, the significant limitation was the model's knowledge of current events (ChatGPT 3.5 was trained on data through September 2021). Since then, ChatGPT has partnered with Microsoft, and many have been using Bing Chat as a stopgap (and, more recently, Google Bard). Now ChatGPT 4.0 includes the ability to use web plugins; up to 3 at a time can be used in a single chat. This allows you to browse with Bing or use plugins; either enables web search to bring more current information into ChatGPT. This is a giant leap forward for the platform, making most knowledge workers’ tasks more efficient with AI.

Unfortunately, the plugin search experience is pretty terrible as I cannot find an easy-to-list directory other than clicking on the Plugins Beta and browsing through the Plugin Store in the ChatGPT Interface. Here are some of the plugins you may find helpful (I like LinkReader and WebPilot, but I am sure there are plenty of good options).

• BizToc: Provides the latest business news updates, market trends, and insightful analysis.

• ChatwithPDF: Helps query PDF documents and extract specific information.

• KeyMate: Allows web searches using a custom search engine.

• Link Reader: Offers features like article summaries, reading time estimates, and read-later options.

• Metaphor: Provides access to the internet’s highest quality content through a powerful neural search engine.

• ScholarAI: Academic paper summaries, research trend analysis, and even assistance with your research work.

• Web Pilot: Browse & QA Webpage/PDF/Data. Generate articles from one or more URLs.More plugins are being added daily, but the ability to interact with the latest data from the web via ChatGPT is a vast improvement for chatbots. If you aren’t using them, I highly recommend using ChatGPT Plus so you are guaranteed access.

What I Read this Week

• Unlocking the Power of LLMs with LangChain: A Beginner’s Guide - Coding the Smart Way

• ChatGPT took their jobs. Now they walk dogs and fix air conditioners - Washington Post

• Yum CEO: ‘AI is going to change how we run our business’ - National Restaurant News

• How the rise of AI is putting flat management structures to the test - Raconteur

• Google, Nvidia and Microsoft Offer What VCs Can’t - The Information (The Information is a paid publication, but I pay for it and love it. If you want to try it out, this link will provide 25% off a subscription).

What I Listened to

Some of the more interesting podcasts on AI I listened to this week.

AI Tools for Enterprises and Business Users

These tools are all very much aimed at creating enterprise and business applications, and most require e of technical prowess.

• Dora - Just like my previous recommendation of 10Web, Dora promises to create websites from a limited amount of text. It’s garnered some attention on Product Hunt, and I am adding it to my one-to-watch list.

• Poe (by Quora) - Like ChatGPT, Poe offers a chatbot interface but lets you choose between ChatGPT, Claude, and soon-to-be-added other models.

• ContentIn - Write better LinkedIn content 10x Faster with AI assistance.

• Sivi - Generate visual designs quickly from a prompt.

Midjourney Prompt for Newsletter Header Image

For every issue of the Artificially Intelligent Enterprise, I include the Midjourney prompt I used to create the feature image. Here’s this week’s prompt.