Art Is in the Eye of the Computer
It turns out that I’ve been writing about Generative AI without even realizing there was something called Generative AI, such as articles about the robot artist Ai-Da, the AI image creator DALL-E, or patent protection for AI inventors. Generative AI refers to AI that strives not just to process and synthesize data but to actually be creative. It’s starting to both become more widespread and to attract serious attention from investors.
James Currier of investment firm NFX sees “Generative Tech” as the next big thing: “If crypto hadn’t happened, we’d probably be calling THIS Web3.” He distinguishes Generative AI from Generative Tech as:
Some have called it “Generative AI,” but AI is only half of the equation. AI models are the enabling base layers of the stack. The top layers will be thousands of applications. Generative Tech is about what will actually touch us — what you can do with AI as a partner.
He predicts Generative Tech will generate “trillions of dollars of value.” I’m hoping that healthcare is paying attention.
Let’s start with OpenAI and its DALL-E 2. DALL-E 2 got much attention earlier in the year with its startlingly unique images, and now is more broadly available, with more than 1.5 million users. The Wall Street Journal calls its images “amazing — and terrifying.”
OpenAI is overseen by a non-profit company, and its mission is “to ensure that artificial general intelligence benefits all humanity,” presumably meant to draw the distinction from AI developed by for-profit companies (such as DeepMind, which is owned by Alphabet). Its charter explicitly states that it seeks that AI/AGI “is used for the benefit of all, and to avoid enabling uses of AI or AGI that harm humanity or unduly concentrate power.” Its “primary fiduciary duty is to humanity.”
Microsoft invested $1b in 2019 (in return for OpenAI using Azure as its cloud partner and giving Microsoft priority in brining new technologies to market), and The Wall Street Journal now reports that the companies are in “advanced talks” for a new round of funding. OpenAI is valued at nearly $20b.
Then there’s Stability AI, which just announced a $101 million funding round that values the company at $1b. It bills itself as “the world’s first community-driven, open-source artificial intelligence (AI) company,” with a slogan “AI by the people, for the people.” Emad Mostaque, founder and CEO, states:
Stability AI puts the power back into the hands of developer communities and opens the door for ground-breaking new applications. An independent entity in this space supporting these communities can create real value and change.
Its competitors to DALL-E are Stable Diffusion, released in August, “a powerful, free and open-source text-to-image generator” that already has been licensed by 200,000 developers, and DreamStudio, its consumer-facing image product that has a million registered users.
A New York Times article noted that Stable Diffusion has limited safety filters, which has made it popular among artists and has led to some, shall we say, objectionable images. Mr. Mostaque is undeterred, telling the NYT: “We trust people, and we trust the community, as opposed to having a centralized, unelected entity controlling the most powerful technology in the world.”
He made a similar point to TechCrunch: “Nobody has any voting rights except our employees — no billionaires, big funds, governments or anyone else with control of the company or the communities we support. We’re completely independent.”
Still another Generative AI company, Jasper, also scored a funding round last week, with the $125 million round valuing it at $1.5b. Jasper bills itself as an “AI Content Platform,” including both generating text and text-to-images. Interestingly, it uses OpenAI’s GPT-3 to power the platform.
Jasper CEO Dave Rogenmoser says: “Generative AI represents a major breakthrough in creative potential, but it’s still inaccessible and intimidating to many. Jasper is working to bring AI to the masses and teach people how to leverage it responsibly so that businesses and individuals can better convey their ideas.”
I’d be remiss if I didn’t at least mention Anthropic, which has raised close to $800 million, or MidJourney, which boosts 3 million users taking advantage of image generator on its Discord server. Anthropic is set up as a public benefit corporation, and is “working to build reliable, interpretable, and steerable AI systems,” while MidJourney describes itself as “a small self-funded team focused on design, human infrastructure, and AI.” I’m sure there are others.
Joanna Stern, writing in the WSJ, marvels: “The stuff once found in AI research labs is now making it into our homes and offices…For decades, we’ve been hearing AI is going to change how we interact with computers and the world. These tools may be the first time most of us recognize it in action.”
There’s already much concern about the “black box” of AI — we may not know how or why AI comes up with things — and the implicit biases that may be built it (e.g., most healthcare data sets will include the results of the inequities in our current healthcare system). Stability AI’s Mr. Mostaque told NYT: “Ultimately, transparency, not top-down control, is what will keep generative A.I. from becoming a dangerous force,” and I hope he’s right — but I’m not sure he is.
I love the idea of “democratizing” AI, putting it in our homes and offices. I like the idea that some of this is happening through non-profits, like OpenAI. I’m highly intrigued that some of it is open source, like Stability AI. And I’m wondering what the impetus in healthcare will be to bring it to our care and to our daily lives.
If Mr. Currier is remotely right that Generative Tech will unleash trillions of dollars in value, healthcare is not going to be untouched. I’d love to see a collection of our healthcare giants — health systems, health plans, pharma, etc. — pool their data for use by a non-profit focused on AI for healthcare. I’d like that AI not just be better supportive tools for clinicians, but also to be creative, up to and including “AI physicians,” whatever they may be. And I think it’d be cool if much of this work could be open source and aimed at the masses.
Mr. Currier predicts: “In the next 10 years, we will expect software to collaborate with us. It will be the new normal. Steve Jobs said in 1980 that the Apple personal computer was a bicycle for the human mind. You might say that Generative Tech is a rocketship for the human mind.”
Buckle up: it’s going to be a bumpy ride.