Upcoming Events
Tue, Mar 12th: AI Open Demo Night - SF March
Wed, Mar 20th: 🧠 GenAI Collective x Daily 🗣️ AI & Voice Summit
Thur, Mar 21st: 🧠 GenAI Collective 🧠 Marin Startup & Scaleup Nerdworking
Wed, Mar 27th: 🧠 GenAI Collective x Girls in Tech 🙍♀️ Diversity and Equity in Tech
🗓️ Hungry for even more AI events? Check out SF IRL, MLOps SF, or Cerebral Valley’s spreadsheet!
If you didn’t get a chance to listen to our most recent episode of the Collective Intelligence Community Podcast, go listen to our very own Thomas Joshi’s interview with Slater Stich, Partner at Bain Capital Ventures, where they explore everything from the current state of the modern data stack to operationalizing the next generation of AI agents to implications of AI on emerging infrastructure startups!
Welcome to the 12th edition of the GenAI Collective Newsletter! We're thrilled to celebrate 6 months of connecting with our amazing community. We're incredibly grateful for your continued support and look forward to continuing to learn, grow, and build together in the months to come. 📖
How LLMs Will Disrupt Online Advertising?
It’s no secret that the rise of foundation models will have a significant impact on Google’s core business model. In the early 2000s, the rise of the consumer internet and the creation of the PageRank algorithm democratized access to relevant information, but now large language models are democratizing synthesis. As GPT-4 and other foundation models proved they could reach (and even surpass) human benchmarks on text summarization, sentiment analysis, and semantic reasoning, users began to question the need and efficiency of the traditional search engine workflow. Why should a user sift through hundreds of relevant links and thousands of pages of content to find an answer when a language model can provide the same synthesis in seconds? At the same time, consumers are tired of distracting, targeted ads; they want answers and actions. The jury is out on what the future of search will look like in the next 5 years, but it’s safe to say the output will not be 50 blue links and annoying ad banners.
The Shift from Ads to Data Monetization
That being said, there is never a free lunch and the internet is no exception. Its economic engine is fueled by advertising revenue, where user data (cookies) and targeted ads are exchanged for “free access.” This model, while wildly profitable for a select group of public companies (experts believe Google’s search business maintains 75%+ operating margins), has been criticized for privacy violations and an increasingly poor, ad-saturated user experience. Regulations continue to grow around data privacy and large tech incumbents that rely on ad revenue are struggling to keep up. Amazon, Meta, Google, and TikTok have all been hit with hundreds of millions of dollars in fines for privacy violations, and consumer trust is rapidly declining. Consumers see LLM-powered search as a viable alternative with a recent Gartner survey finding that 79% of consumers are excited to use AI-powered search and over 70% trust the results. Further, a recent study (source below) with 120 participants found that consumers are already showing signs that they prefer an LLM-based search experience.
LLM-powered search introduces a paradigm shift in the internet’s business model. The performance of these models rely on an ever-increasing feed of high-quality training data. Although many foundation model providers have not released the number of training tokens used for their most recent models, the number is understood to be much larger than previous models.
As raw data becomes the primary asset for foundation model providers, internet companies, creators, and publishers can monetize on the quality of their content. In Reddit’s S-1 filing, the company disclosed that 10% of their total revenues, or $60M, come from selling their data to train large language models. They also stated that “We expect our growing data advantage and intellectual property to continue to be a key element in the training of future LLMs.” This marks a significant departure from relying solely on ad revenue and showcases the potential of data as a direct revenue source.
Redefining Business Strategies
This new focus on data monetization will force businesses to reimagine their monetization strategy. Gartner predicts brands’ organic search traffic will decrease by 50% or more as consumers embrace LLM-powered search. Internet companies will need to prioritize the collection and curation of high-quality data that can improve LLM performance over generating clickbait content for ad impressions. This strategic pivot could significantly impact how content is created and managed online, emphasizing the value of data richness and uniqueness.
Additional examples include:
StackOverflow & Auttomatic: StackOverflow signed a deal to sell data to Google for LLM training and Auttomatic, the company that owns Tumblr, signed deals with Midjourney and OpenAI to train models on their data.
xAI: According to Reuters, Elon Musk is looking to raise up to $6B for his new AI startup xAI and believes X’s vast stores of original content will be a key value driver and differentiator for the business. X’s generative agent Grok has already shown impressive results in recent benchmarking exercises.
GPT-4 and Content Creation Platforms: Platforms that offer content creation services using GPT-4 showcase another facet of this transformation. By generating unique content, these platforms can create valuable datasets for further LLM training, simultaneously offering innovative services and creating data for sale.
Language Learning Applications: Duolingo, a language learning app, utilizes LLMs to enhance its teaching algorithms and user experience. The data generated from user interactions could be repurposed to train more sophisticated LLMs, presenting an additional revenue stream while improving product offerings.
The potential for reducing the reliance on advertising could usher in a more user-centric internet. If companies can sustain themselves through data sales, the need for online advertising will decrease, leading to an enhanced online experience. This could encourage a more thoughtful design of the search experience, prioritizing user experience and quality over impressions and engagement. How will this change the business model of traditional search incumbents? Which companies are set to profit the most from data sales? Are we going to see a rise of data marketplaces? Would love to see you all at our next event to continue the discussion!
With that, I hope you enjoyed the newsletter and would love to hear your perspective on how you see LLM-powered search disrupting the ecosystem. If you have any ideas or insights you would like to share, hit up Eric on the community Slack or reach out via email at eric@gaicollective.com for a feature in our next newsletter!
Events Recap
GenAI 🤝Product + Design Workshops
In February, we hosted back-to-back GenAI workshops for the Product and Design communities, collaborating with Ansh Mehra (world’s top AI YouTube content creator) and Grace Ling of Design Buddies (world’s largest design community) to put on a series of info-rich events!
We also partnered with speakers from Product Hunt, Kraftful, Cofactory, AWS, and Meta to put on a set of panels, where we got to hear about the integration of GenAI into the crafts of Product and Design from top practitioners in the industry, as well as implications for those fields in the upcoming years!
The crowd reception was incredible - people appreciated the practical tips from the workshop, and the forward-looking nature of the panels. Special shoutout to Cloudflare to hosting both events - looking forward to many more!
About Eric Fett
Eric joined The GenAI Collective in early September to lead the development of the newsletter. He is currently an investor at NGP Capital where he focuses on Series A/B investments across enterprise AI, cybersecurity, industrial, and space technology. He’s passionate about working with early-stage visionaries on their quest to create a better future. When not working, you can find him on a soccer field or at a sushi bar! 🍣