A research and deployment company developing realistic, versatile AI audio software for creators and publishers. A collaborative software-building platform and an AI-powered code-generating tool. But based on the early data we have for generative AI, combined with our experience with earlier AI/ML companies, our intuition is the following. Interestingly, the gains offered by the microchip and the Internet were also about 3-4 orders of magnitude. In the image generation category, the “barrier to launch” an app is fairly low, thanks to third-party APIs.
Advances in LLM steering also have the potential to unlock new possibilities in sensitive consumer applications where users expect tailored and accurate responses. Some have pointed out that LLMs are poised to unseat entrenched consumer applications like search, but we likely need better steering to improve model outputs and build user trust before this becomes a real possibility. Improved steering becomes especially important in enterprise companies where the consequences of unpredictable behavior can be costly. Improved steering will also pave the way for broader adoption in other industries with higher accuracy and reliability requirements, like advertising, where the stakes of ad placement are high.
Regardless of where defensibility comes from and who ultimately captures market value, the consumer will ultimately be the biggest winner. A paper back in 2019 found that consumers value “free” products in shockingly big dollar terms, estimating a willingness to pay as high as $17.5K for search engines, $8.4K for email, and $1.2K for streaming services. And if software history tells us anything about innovation, it’s that great entrepreneurs will always find ways to build important, durable companies in each new technological era.
And the same concerns of “hate speech” (and its mathematical counterpart, “algorithmic bias”) and “misinformation” are being directly transferred from the social media context to the new frontier of “AI alignment”. Historically, every new technology that matters, from electric lighting to automobiles to radio to the Internet, has sparked a moral panic – a social contagion that convinces people the new technology is going to destroy the world, or society, or both. The fine folks at Pessimists Archive have documented these technology-driven moral panics over the decades; their history makes the pattern vividly clear. It’s worth noting that dynamically generated worlds on their own are not enough to make a good game, as evidenced by the critical reviews of No Man’s Sky which launched with over 18 quintillion procedurally generated planets. The promise of dynamic worlds lies in its combination with other game systems – personalization, generative agents, etc – to unlock novel forms of story-telling.
DirectMusic was never widely adapted, due largely to the difficulty of composing in the format. Only a few games, like Monolith’s No One Lives Forever, created truly interactive scores. It’s important, since it can help set the emotional tone just as it does in film or television, but since games can last for hundreds or even thousands of hours, it can quickly become repetitive or annoying. Also, due to the interactive nature of games, it can be hard for the music to precisely match what’s happening on screen at any given time. We’ve seen a few initiatives in the space, like Promethean, MLXAR, or Meta’s Builder Bot, and think it’s only a matter of time before generative techniques largely replace procedural techniques.
As with any emerging technology, generative AI has been met with some criticism. Though some of this criticism does reflect current limits of LLMs’ current capabilities, we see these roadblocks not as fundamental flaws in the technology, but as opportunities Yakov Livshits for further innovation. In addition, we see significant potential in enterprise-oriented applications for internal search. Most companies now use a number of communication apps and databases, such as Gmail, Slack, Drive, Asana, and more.
Note – there are many challenges to be solved still before we see a fully generative version of the Sims. LLMs have inherent biases in their training data that could be reflected in agent behavior. The cost of running scaled simulations in the cloud for a 24/7 live service game may not be financially feasible – operating 25 agents over 2 days cost the research team thousands of dollars in compute.
Founder of the DevEducation project
A prolific businessman and investor, and the founder of several large companies in Israel, the USA and the UAE, Yakov’s corporation comprises over 2,000 employees all over the world. He graduated from the University of Oxford in the UK and Technion in Israel, before moving on to study complex systems science at NECSI in the USA. Yakov has a Masters in Software Development.
It is also important to note that while incumbents appear to be adapting faster, the shift to AI is also playing out faster than the shift to cloud and SaaS. It took ~5 years for SaaS apps to really get going once cloud infrastructure was in place. But in just 6 months, ChatGPT reached 300M users and launched plug-ins and APIs for developers to build on top of GPT. As a result, the talent and effort needed to ship an AI application are far less than what was needed to adapt an on-prem product for cloud—and it will only continue to get easier as the LLM ecosystem and off-the-shelf tooling further mature. A second group of products may tolerate probabilistic outputs but don’t necessarily benefit from the non-deterministic nature of the platform. For example, think about products that synthesize existing content, where variability in how the synthesis is presented is generally fine so long as the gist of the synthesis is accurate.
The issue with AI historically is not that it doesn’t work—it has long produced mind-bending results—but rather that it’s been resistant to building attractive pure-play business models in private markets. Looking at the fundamentals, it’s not hard to see why getting great economics from AI has been tough for startups. 90% of companies on the list are already monetizing, nearly all of them via a subscription model.
My view is that the idea that AI will decide to literally kill humanity is a profound category error. AI is not a living being that has been primed by billions of years of evolution to participate in the battle for the survival of the fittest, as animals are, and as we are. It is math – code – computers, built by people, owned by people, used by people, controlled by people. The idea that it will at some point develop a mind of its own and decide that it has motivations that lead it to try to kill us is a superstitious handwave. The presumed evolutionary purpose of this mythology is to motivate us to seriously consider potential risks of new technologies – fire, after all, can indeed be used to burn down entire cities.
It’s even possible to one day imagine an entire personalized game, created just for the player, based on exactly what the player wants. This has been in science fiction for a long time—like the “AI Mind Game” in Ender’s Game, or the holodeck in Star Trek. But with the tools described in this blog post advancing as quickly as they are, it’s not hard to imagine this reality is just around the corner. At this point there are hundreds of companies building general purpose chatbots, many of them powered by the GPT-3 like language models.
Now we’re seeing a number of companies trying to create AI generated music, such as Soundful, Musico, Harmonai, Infinite Album, and Aiva. And while some tools today, like Jukebox by Open AI, are highly computationally intensive and can’t run in real-time, the majority can run in real-time once the initial model is built. Creating great animation is one of the most time consuming, expensive, and skillful parts of the game creation process. One way to reduce the cost, and to create more realistic animation, is to use motion capture, in which you put an actor or dancer in a motion capture suit and record them moving in a specially instrumented motion capture stage. We’re seeing several different startups going after each stage of this 3D asset creation process, including model creation, character animation, and level building.
There are also significant concerns over how to compensate the original writers, artists, and creators behind training data. The challenge is that most AI models today have been trained on public data from the Internet, much of which is copyrighted work. In some cases, users have even been able to recreate an artist’s exact style using generative models.
But just as fire was also the foundation of modern civilization as used to keep us warm and safe in a cold and hostile world, this mythology ignores the far greater upside of most – all? – new technologies, and in practice inflames destructive emotion rather than reasoned analysis. Just because premodern man freaked out like this doesn’t mean we have to; we can apply rationality instead.
Products like AdCreative and Pencil can produce marketing collateral for email or social media, while Frase or Writesonic can write SEO-optimized product descriptions. Eventually, we expect users will be able to create an entire ecommerce store—and the materials to market it—by simply describing their desired aesthetic and clicking a button. Not only will AI propel the creation of more games, but it will advance a new type of game that is more dynamic and personalized to the preferences of each gamer. We’ve already seen some early examples of this with text-based games like AI Dungeon and Hidden Door. Imagine entering a game and being able to design a sophisticated custom avatar with just a few sentences. Eventually, this may expand to entire virtual worlds you can create from scratch.