The Sound of Silence
5 multi-million-dollar generative AI businesses someone needs to build right now
In my role as general partner of Project Voice Capital Partners (and the PVCP Rolling Fund), we’re seeing a wide variety of young companies and talented entrepreneurs embracing generative AI.
Here’s five different generative AI companies/tools that need to be built right now.
Never say I didn’t give you anything.
AI tool that creates compilation reaction videos
In 2016, metal rockers Disturbed took the stage on Conan to perform their cover of Simon and Garfunkel’s The Sound of Silence.
Little did anyone know this magnificent performance would go on to become one of the most viewed music performances in the history of the internet, and certainly the largest in the history of Conan O’Brien’s show.
The performance is here:
Now, as people will do, they film “reaction videos” which are nothing more than videos of random people watching the clip, ostensibly for the first time, and reacting to it in real-time in front of the camera.
Beyond this, there are even more videos in which creators compile all of these various reactions, particularly the better ones, and create compilation reaction videos. For Disturbed’s performance above, the biggest compilation reaction video has over half a million views - well past the zone of being a profitable endeavor for the creator.
If you told me generative AI created this, I’d easily believe you - it could be tightened in places for sure - but it’s good enough. And generative AI easily could do videos like this for many, many other performances and types of videos that draw reactions.
YouTube’s economics would make whoever creates this an instant multi-millionaire.
Consulting agency that guides enterprise through AI regulatory compliance
The European Union has already rolled out proposed high-level legislation around artificial intelligence, called The AI Act. (Credit to Christoph Borner of Cyara for alerting me to this.)
Like GDPR, this legislation might have legs and go on to be adopted by many companies in other geographies, like it has by many in the United States. But even if it doesn’t, the time where enterprise will need to navigate governmental regulation is almost here.
Consider the difficulty large companies would have in answering basic questions about AI, such as “describe the percentage of your external marketing that includes AI?” or “describe the percentage of your internal operations that involves AI?” Almost no company today would have any idea how to compute that. Start there.
AI tool to create musical mashups
There’s nothing better than a musical mashup, generally two distinct songs combined into one seamless, singular “mashed up” tune.
When these are done well, they get millions of views on YouTube alone. Some good examples…and a warning that you won’t ever hear any of these songs the same way again…
Once someone comes up with a generative AI tool that creates these mashups, after assessing and perhaps match-making compatible songs and having clever algorithms behind how to construct mashups creatively, it will print money on YouTube and perhaps in other ecosystems like Spotify.
Generative AI comparative analytics
While the media fixates on how to compare ChatGPT’s output to that of other humans, there’s a larger and more immediate target: figuring out how to compare ChatGPT’s output to that of other text-based generative AI solutions instead.
We need someone with a mathematics background to conjure up a scoring algorithm that will assess the ‘fit’ of a generative AI text answer to the original query, and assess which of two or more answers (across generative AI frameworks) is superior to the others, along numerous metrics synthesized into one core quantitative score.
Think of it as standardized testing for AIs, rather than humans, used to reduce the complexity of comparing AIs with each other.
“Create A New Episode”
For TV shows, book series, podcast series, or other types of serial fiction that have sufficient amounts of stories (maybe at least two or three seasons’ worth), there should be a generative AI solution that can, on a whim, produce a brand new story from that universe.
For example, I would gladly pay $5-$10 for an AI-generated brand new episode of my favorite TV show of all time, Star Trek: The Next Generation.
I would expect such a thing to be in a written format, rather than in some sort of AI-generated avatar-oriented video with synthetic voices. I would likely prefer it to be just words, even though the original material was obviously a TV show.
But with enough training data to draw from, from multiple existing seasons, an AI that could do this would be able to pay a licensing fee back to the rights holders, creating a new revenue stream for them, while giving fans an incredible new entertainment option at their disposal.
Somebody start making this stuff!
We’ll have a Project Voice 2023: The World of Conversational AI preview newsletter in early April, and then will take a break over the summer before returning in the fall alongside Season 9 of This Week In Voice, the podcast.
We first published a list of 100 attending companies a few weeks ago, and today published this list of 250 attending companies, with new attendees registering daily. We’ll likely publish one more list before the conference.
10 Louder, 1904 Labs, 3CLogic
A Public Space, AARP, Acapela, Achronix, Actionable Marketing Guide, Aging and Health Technology Watch, AI Data Innovations, Ai4, AKFI, Amazon, Amelia, Appen, Applied Brain Research, Artificial Solutions, AssemblyAI, Asurion, AT&T, Attn Live, Ava, Avaamo, AWS, Ayden Corporate Development Group
Babbly, Behavioral Signals, Besnoy Law, Blue Check Skill, Blue Cross Blue Shield, BNH AI, Booz Allen Hamilton, Brighten AI, Brilliant
Canary Speech, Carhartt, Cedar Mountain ConneXus, Cerence, CereProc, Chick-Fil-A, CO LAB, Cobalt Speech, Colliers, Comcast, Connected Vehicle Trade Association, conversationHEALTH, ConverseNow, Creative House, Creative Virtual, Creci Ventures, Credera, Crunchbase, CVS Health, CX Forums, CXO2, Cyara, Cyrano AI
D3 Institute at Harvard Business School, Dashbot, DataForce, Datasaur AI, Datch, Deepgram, Defined.ai, Deloitte, DignifAI, Dream Builder, Druid AI
e2f, Educated Industries, Elavon, Elevance Health, Encounter AI, EPB, EXL, EY
Fairfax Software, Fandango, Fiserv, Flipkart, flowy.live, Forbes, Frost & Sullivan, Future of Privacy Forum, Futurewei Technologies, Inc.
G2, General Motors, Genesys, Georgia State University, GlobalLogic, Got It AI, Grammarly
Harvard Business School, HCA, HEAD acoustics, Holon Institute of Technology, HomeServe, Hyro AI
IBM, IDVerifact, Immaculate Conception Basilica, In The Room, Inbenta, Indiana University, Innodata Inc., Instacart, Intel, Intelligent Health Association, Interactions, Intis Telecom (it.com), Invoca, iScribeHealth
Kaleidoscope, Kindspace, Kinetic Communications, KomBea, Korn Ferry
LAA, Inc., LBMC, Lenovo, Lexmark, LifeScore Music, LillyPad AI, Linux Foundation, Livinguard, LMO Communications LLC, Logitech, LXT
Marlinspike, Mars Petcare, ME4U AI, Microsoft, Moveworks, Mozilla, MSH, Myrtle Software Ltd
Naris Communications LLC, Nation's Restaurant News, National Retail Concept Partners, Native Voice, NBC News, New York Times, Next Insurance, Nissan Motor Corporation, NVIDIA, nVoq
Olympus, OneReach.ai, OpenDialog AI, Orbita, PandoLogic
PayTalk, Penguin Random House, Perfect Bound, Planisware, Play HT, PolyAI, Print Media Centr, Private AI, Project Voice Capital Partners, Prudential, Publicis Sapient, PwC, PYMNTS.com
Raddle, ReadSpeaker, Reel2Media, Regions Bank, RelationalAI, Resemble AI, Resideo, Restaurant Business, Restaurant Technology Network, RetailWire, Riva Health
Sanas AI, SAP, SapientX, Save A Lot, Schneider Electric, Sensory, ServiceLink, ServiceNow, Shaip, Shell, Sheridan College, Siemens, Singularity University, Slalom, Slang Labs, Slate, SmartListener.ai, Social Evolution, SolomonEdwards, SoundHound, Southern Company, SparkleHaze, Speech Accessibility Project, Speechly, Speechmatics, Stanford University, Sunbelt Rentals, Symbl.ai
Tableau, Talkwalker, TEKsystems, The Atlantic, The IVR Voice.com, TransUnion, Trigran
UC Berkeley, UC Berkeley - Haas School of Business, UCB, Uniphore, United Ability, University of Cincinnati, US Bank, US Department of Energy, US Department of Veterans Affairs, Valence Vibrations, Vanderbilt University, Verbit, Verint, Veritone, Virtual News Center, Voicebot AI, Voiceflow, VoicePlug, Voices, VoiceXD, Volvo, Vozy Inc, VUI Agency, VUX World
Wall Street Journal, Walmart, Wavestone, WealthVoice, WellSaid Labs, Welocalize, WESTCARE, WillowTree, Women in Data, Wysa, Wysdom
YUM Brands
ZKH, Zoom
The main gift received from attending Project Voice 2023: The World of Conversational AI is the ability to alter your trajectory.
Whether a buyer looking to invest in conversational AI in the wake of ChatGPT's arrival, a vendor looking to connect with customers and partners, media looking to cover the Ethics & Integrity Charter on LLM-Based AI that will be released at the conference, or investors looking to deploy capital in this new space, the gift is networking of the highest order.
But hey, we'll give every attendee a fancy new pair of socks too.
Thank you to Techmeme, as usual, for their partnership in promoting Project Voice 2023: The World of Conversational AI.