By David Edelsohn, ROBO Global Venture Advisor & Senior Technical Staff Member at IBM Research
A leaked internal report from Google reveals that even pioneering companies developing Generative AI (GenAI) to create new solutions are questioning their role in the future. In this case, the unease comes not from fear of AI’s ability to replace human workers, but from whether a combination of GenAI open source design and ease of use can eliminate any competitive price for solution providers.
The technology and algorithms underlying GenAI models are complex, so coming up with the basic concept was difficult and time-consuming. But with the heavy lifting done, creating models and applying them to various use cases is now only limited to your imagination. This reality presents a significant challenge for Google and all other GenAI solution vendors. As the report states, “We have no moat, and neither does OpenAI.”
Welcome to the latest technological commodity revolution. Hardly a day goes by without a startup announcing the launch of a new market-specific GenAI product. The potential for new applications using GenAI is endless, and the friction and barriers to creating new applications have already disappeared. For GenAI companies around the world, this rapid increase in activity raises the question: if the best of everything can be easily duplicated or created, will there be such a thing as a “competitive moat” in the future? As daunting as this threat may seem to GenAI providers and traditional enterprise service organizations, previous technology revolutions can shed light on what we can expect moving forward.
History may not repeat itself, but it sure rhymes. In the 1980s, the personal computer revolution opened the floodgates of software development, created some of today’s leaders in technology companies, and made Bill Gates and Steve Jobs famous. Seemingly overnight, software development moved from being the exclusive domain of an elite few working for large corporations to being open to anyone with enough programming skills to create a software application. Soon after, access to powerful systems, virtualization, and the Internet created further commoditization across the industry, allowing a startup company to be created in the cloud with little or no physical infrastructure. It was a whole new world.
Indeed, during each of these eras of change, industry leaders may have felt that their “competitive moats” had evaporated. Of course, we know the opposite was true. Early adopters and adopters were able to take advantage of the environment and excel. They have built technology ecosystems and scalable business models. They have created massive network effects to rapidly increase the value of their solutions and focus on delivering superior user experiences. And the momentum created by companies like Apple, Microsoft, and Amazon has given them a valuable head start, allowing them to reduce the friction of scaling and focus on another revenue generator: data management. These early winners remain among today’s technology leaders.
Will GenAI present a different scenario? In this case, the level of commoditization seems even more extreme. Threatened by AI’s ability to deliver results that were always thought to be the preserve of the human brain, content creators and creatives are doing everything they can to protect their livelihoods. When GenAI created the song “Heart on My Sleeve” expertly spoofing top artists Drake and the Weeknd swept the internet earlier this year, proving to the world that artificial intelligence can do more than just that write song in the style of the artist, but even to perform in the style of an artist. The music world immediately went into crisis mode. Hollywood is also feeling the pain, with AI limitations a key talking point at the moment Writers Guild of America (WGA) strike.. Other unions are obliged to follow suit. They have a right to be concerned. Multimodal generative AI– which includes text, images, video, voice and 3D – is still in its infancy, but with the rapid progress and pace of change, it seems imminent that AI will soon be able to handle every element of project development and execution – from selecting lucrative audiences and shaping story, writing the script, planning the shot list and “casting” the synthesized actors for each role. It will then be used to render scenes, edit the film and write an original score – all without the need for the layers of executives or teams of creative professionals that traditionally bring a story to life.
Brands representing famous actors and big franchises like Star Trek, Star Wars and all things Disney are already working to negotiate commissions to help protect their intellectual property. We have moved from a world where influence was controlled by people to a world where distribution channels like Apple, Netflix and Spotify rule. Furthermore, we can see savvy entrepreneurs creating completely fictional virtual characters to appeal to the next generation, who are looking for brands that reflect their own experiences and identities.
The challenge is accelerating every day. GenAI capabilities leapfrog past results at lightning speed, rapidly democratizing open-source tools like GPT, LoRA, QLoRAand more Large Language Models (LLM). With the massive improvement in training performance enabled by GenAI, the next wave will not be larger models, but greater optimization and specialization to deliver new information in near real-time and to produce customized models for individual users and devices. And as low-cost customized models prevail, they will enable a streamlined path to much greater data privacy, which could lead to deplatforming and ultimately reduced reliance on big tech names.
If we see a decline in the dominance of the big tech names, tomorrow’s winners will be the companies that build their strategies around GenAI adoption. Their leaders will think like previous generations of technology visionaries—Nadella, Gates, Bezos, Musk, and Jobs—who recognized technology’s potential and delivered on that promise. In an environment where technology is open and democratized, those who are able to identify areas to dominate will rise to the top.
Tea ROBO Global Artificial Intelligence Index (ticker: THNQ) invests in companies across the AI ecosystem that are poised to become leaders in this dynamic new environment. It seems likely that the early explosion of GenAI startups will eventually consolidate, as those who get in on the act will use their initial war chests to acquire worthy players to create their own GenAI platform, or at least operate as a conglomerate. Some of the companies in the THNQ ETF may participate in these M&As, giving investors access to the results of this growth. Others are key parts of the infrastructure that drives the training and deployment of these AI technologies behind the scenes, and are more immune to which models ultimately become the most successful.
Google’s success mantra was “Dominate, then monetize”. Thanks to GenAI’s open source design and ease of use, Google and many other big tech players may not have the luxury of time to implement this approach. If Google’s leaked memo is correct, their competitive moat is already gone. The result: the door is wide open for lesser-known companies that can quickly build a great “poker hand” of related GenAI capabilities and provide that platform ahead of the pack. May the best player win!
