Launchable: Generative Apps - building the next big wave of AI startups

By Jay Bartot, Partner • July 24, 2023

Wow!  What an extraordinary era of technological advancement we're living in! Over the past year or so, the buzz around "Generative AI" has been nearly impossible to miss. This transformative technology is shaping up to be as revolutionary as the PC, the Internet, and the iPhone. Years from now, we'll likely wonder how we ever managed without it.

For those of us who thrive on innovation, these are thrilling times. The chance to be on the ground floor, shape the future of technology and make a lasting impact is what motivates us. At MVL, our team has spent many years at the forefront of enterprise AI and machine learning. The current wave of developments in Generative AI has us more excited than ever to do what we love - working with and helping budding entrepreneurs with a vision and passion create and build data-driven AI companies! 

We're excited to announce our upcoming event, Launchable: Generative Apps, happening this October. Tailored for Gen AI entrepreneurs, this event promotes the inception of groundbreaking, scalable startups. It's the ideal venue for AI entrepreneurs to connect, share ideas, and fuel their passion for creating Gen AI startups with defensible moats. Participants will have the opportunity to launch their vertical, Gen AI startup, with $250,000 in pre-seed funding from Madrona Venture Labs. Work and learn from fellow entrepreneurs, foundation model experts, and VCs.

A plethora of greenfield opportunities!

Maybe you have a concept or idea you’ve been noodling for a while. Or perhaps you’re intrigued by the possibilities but don’t know where to start. Maybe I can help. 

Over the past year, we’ve been delving deep into GenAI technologies, running experiments, and building prototype products and POCs. I've been sharing “OMG” moments with friends and colleagues as I experiment with the quickly evolving Gen AI stack, including the latest Large Language Model (LLM) models and OSS toolkits such as langchain, llama-index, and auto-gpt. Upon reflecting on these moments, it struck me that each represents a vast realm ripe for creating innovative products and companies. Let's consider a few:

• I was astounded when an LLM accurately dissected a random Python function I fetched from Github and proposed a series of enhancements, delivering a more refined piece of code. The idea of a product manager at a software company inquiring about the functionality of a specific feature directly from the source code, bypassing the need for a software engineer, is game-changing. This showcases Gen AI's enormous potential for transforming software and IT development.
• LLMs' ability to decipher jumbled, informally written, jargon-infused eCommerce product reviews is a feat previous generations of NLP technology couldn't master. Or what about the troves of semi- or unstructured public government data and reports that contain scattered breadcrumbs leading to how decisions are made and money is spent? And consider the increasing rate at which we gather and store data just in the past decade. Current estimates suggest we generate approximately 328.77 million terabytes of data daily! An LLM's potential to interpret, connect, and liberate value from these vast, ever-expanding data stores has boundless opportunities.
• With all of the talk about large language models, don’t forget about images and video.  Technologies like Stable Diffusion and products like MidJourney are creating extraordinarily realistic images from text input. Their ability to automatically generate visual narratives that enhance storytelling, transcending my modest artistic skills, is epic. With a virtual visual artist at our side, we'll discover novel ways to communicate, educate, market, and sell products, services, and ideas.
• Automatic summarization capabilities of lengthy content offer new paradigms for creating and consuming information. Given the flood of information that passes through our inboxes, news readers, and social and video feeds, deciding where to focus and devote time is often overwhelming. A system capable of presenting a multi-resolution view of information, starting with a brief summary and then allowing users to iteratively delve deeper into details (akin to zooming into an image), will change how we consume information.
• Using OSS toolkits, I was astounded by the ease of having an enlightening and interactive Q&A conversation with an entire collection of static PDF or Word documents with just a few lines of Python code! This ability to harness dusty, dense text information and make it accessible using LLMs and open-source toolkits is phenomenal.
• While still in its early stages, we've already seen the uncanny ability of an LLM to formulate action plans for various tasks. The concept of agent technologies executing multistep tasks on our behalf and freeing us up for higher-level conceptual work holds immense potential. Whether it's planning a trip or handling financial or administrative tasks that require interacting with multiple applications, many of these everyday tasks, under our guidance, will soon be executed on our behalf.

Even these “OMG” moments represent just a few areas with potential for incredible new transformative vertical SaaS and Enterprise applications. We can’t wait to hear your ideas!

Wanted: AI entrepreneurs and innovators

While Generative AI presents a world of rapidly unfolding possibilities, numerous unanswered questions, issues, and challenges require solutions before the Gen AI wave reaches its full potential. These challenges actually provide more opportunities for entrepreneurs to build new products, platforms, and solutions. For instance, 

• How can GenAI be securely implemented within enterprise environments without compromising information security, compliance measures, corporate policies, etc.? 
• Are there imminent solutions to combat model hallucinations and the integrity of the output? 
• In what scenarios will natural language interfaces empower and accelerate the productivity of information workers? 
• What’s the best way to authenticate the sources of generated content and reduce the spread of disingenuous or harmful narratives?  
• And crucially, how can we significantly reduce cultural biases in the data and models, safeguarding vulnerable members of society from harm?

Join us and be part of this momentous technological revolution. Think where the technology will be in three or five years! LLMs will become more powerful, efficient to train, and more cost-effective. Now is the time to take the plunge and be a part of this major wave.

Apply for Launchable: Generative Apps today!