Much ado about generative AI: the money, the potential, the pitfalls

VCs are drooling over generative AI companies and opening their checkbooks to very large rounds of funding despite the current economic climate. In the second half of 2022, in the US alone, Jasper raised $125 million in Series A financing while Stability AI secured a $100 million round.

To put this in perspective, Cooley’s Q3 Venture Financing Report showed that the pre-money valuation for Series A deals had fallen overall – $58 million in June to $45 million in September (the lowest since July 2021). So to see such high series A rounds in generative AI companies is quite remarkable.

And then there is Microsoft’s multi-year investment in OpenAI, said to be a whopping $10 billion, as ChatGPT is integrated to power the Bing search engine. Companies like Alphabet and Nvidia are also said to be exploring new investments in generative AI.

Some of the reasoning behind this momentum was recently explained A16Z blog post Through Matt Bornstein, Guido AppenzellAnd Martin Casado. As they wrote, “Models such as Stable Diffusion and ChatGPT are setting historical records for user growth, and several applications have reached $100 million in annualized revenue less than a year after launch. Side-by-side comparisons show AI models outperform humans in some tasks by several orders of magnitude.” Simply put, generative AI can be big business.

Generative AI is also seeing significant global appeal, particularly in the APAC region. In fact, Asia Pacific is expected to grow even faster than the US, with a CAGR of more than 35% from 2022-2028. The market is driven by major government initiatives and the widespread adoption of AI-based applications. This will further increase innovation and demand for new offerings around the world as we are more inhibited by imagination than by technology.

But it’s not just the techies who are getting excited about generative AI. Ordinary people are also fascinated. From classrooms to local gyms to dinner parties, people are bringing generative AI into the mainstream by playing with ChatGPT or capturing images with Lensa AI.

Lensa AI, a paid app, is even seeing a lot of interest among consumers. Lensa AI has generated revenue of $16.2 million in 2022, with $8 million in December 2022 alone. The app has been downloaded over 25 million times, with 1.1 million active users as of December 2022. ChatGTP’s numbers are likely to be even more impressive with its paid product. After all, ChatGPT has crossed 1 million users within a week of its launch and continues to captivate users and quickly gain popularity.

So what does it all mean, and where is generative AI headed?

The new creators

One of the most prolific use cases for generative AI is in content and entertainment creation. People like to see what they can put together an AI model – be it with text, speech, static images or even video – that can then be shared with friends, family or the rest of the world. The potential for building new avatar races could be particularly attractive if the metaverse reaches its potential. A new world is opening up.

For example, generative AI can help people without relevant skills or expertise use or create art. Previously, designing, as it is now developed by the layman, required skills such as handling Unity 3D, Unreal Engine, Adobe Creator tools, and so on. Even the most limited compilations required some kind of in-depth knowledge, training, or special equipment.

Today we live in a time where everyone is a creator, but not everyone has the skills to create a great work of art. Generative AI levels the playing field and democratizes art.

Humans don’t just use generative AI to create visual works or produce content, such as a script or blog post. There are also advancements that allow multiple components to be brought together, such as avatars that can be used in videos or chat.

Why would something like this matter other than something fun to show friends or another way to make a school project? Generative AI can be so much bigger than that! It helps people of all ages and in every field to bring their imaginations into the real world.

However, if we’re looking for “big money” business, think about when a casting director tries to find the right actor for a movie role or someone has an idea for a YouTube video that has the potential to get millions of views . It is now possible to generate an “ideal” actor using an avatar tool that can display emotions, change the tone and expression of the voice (think creating voice identities that don’t exist in this world or applying a licensed, distinctive voice, such as Homer Simpson or James Earl Jones), and much more — all before you start looking for a human actor.

Such a use case removes risk and saves time by identifying the right attributes and testing someone’s vision before moving forward with a project. It is much more accurate and efficient. They can see what works – and what doesn’t. Ultimately, making the right decisions from the start can save a project up to several million dollars, depending on its size.

Disadvantages of generative AI

But as with any new or evolving technology, there have been problems. On the one hand, there is the possibility that one’s likeness and views may be co-opted or misinterpreted. Fortunately, advanced technical solutions are rapidly being developed to combat embezzlement. If someone creates an avatar and covers a sensitive topic, an example-based filtering algorithm can be applied, allowing an AI system to understand what their avatar would actually say or how it would respond. These capabilities go far beyond traditional filtering algorithms, which only reject things like swearing.

Creation tools can present the AI ​​system with a sample political issue or article and the user can provide feedback (good or bad). Based on that feedback, the avatar knows what to say or not to say (details of the approach can be found in the research paper User-defined content detection framework).

Similarly, OpenAI recently released a detection tool to differentiate AI writing from writing done by a human. A word of caution is that it is not yet 100% reliable. Errors persist as this is very new territory, but progress is being made. Societal consensus and ethics on the use of such technology are becoming increasingly important as complex copyright, IP, and plagiarism issues continue to emerge, with greater adoption and new use cases emerging daily.

Artists, writers and other creatives have spoken out about how generative AI is a form of cheating, one that devalues ​​real art. New lawsuits are making their way through the courts. A recent lawsuit against Stability AI, image generator startup Midjourney, and online gallery DeviantArt, brought by three working artists, alleges that AI image generators are nothing more than “21st-century collage tools that violate the rights of millions of artists.”

Another lawsuit filed by Getty Images is related to copying images without a license and violating intellectual property. The suits highlight two key issues related to generative AI.

First of all, the definition of Art. What is Art? Who decides what it is and how it should be made? New generative AI creations are still the product of one’s mind and experience; they are still expressions meant to evoke emotions and different reactions. It’s just that the method changes. Instead of creating with brushstrokes, the artist creates with questions and assignments.

Second, who owns what in a product made by generative AI? This is a little stickier, but there’s still plenty of safe space. Images, voices, text, etc. that are properly licensed and/or cited may be used, as well as public domain work. Moreover, generative AI can draw on features of a voiceprint, facial expressions, etc., to build something that doesn’t exist – and never did – in the real world.

Expect many more licensing deals in the future that reflect a changing world integrating generative AI, as well as changing laws related to intellectual property and copyright. This is inevitable as generative AI takes off.

Generative AI is just getting started and there is still so much to discover. The potential is there to open up a whole new creative universe, but only if we ask the right questions and give the right commands in creation. Just as AI systems get smarter, so will the people using them. And with the right tools, the results will amaze us all.

Taesu Kim is the CEO of Neosapience.