As OpenAI pursues a public benefit structure and global expansion, the influence of the AI powerhouse cannot be understated. However, where hundreds of new businesses jump on the AI bandwagon each year to follow in the footsteps of those successful before them, 90% will inevitably fail.
Every few years, a new hype cycle emerges. Venture Capitalists (VCs), under pressure to deliver exponential returns to their limited partners, are constantly searching for the next big thing. In the 90s, it was the ‘dot com’ boom; today, it’s AI.
While a few pioneering companies are genuinely innovating and developing groundbreaking AI technology, a perfect storm is brewing in the background. As the hype cycle intensifies, a multitude of players, both large and small, rush to capitalize on the trend, often without a clear understanding of the underlying technology or its potential applications.
An accelerating trend
Several macro factors have converged to fuel AI’s rapid advancement. The huge decline in the cost of computing power and storage, coupled with ubiquitous internet access and decades of algorithmic research, have created the ideal conditions for AI’s practical application. While academics have long theorized about the potential of AI with vast datasets, the prohibitive cost of infrastructure has hindered their ability to make these theories a reality.
However, a major shift occurred as the cost of supporting AI models became more affordable and accessible. This democratization of AI infrastructure paved the way for deep thinkers to showcase the technology’s capabilities and possibilities. The myriad of potential business applications for AI sparked a wave of interest and investment. For example, Google’s acquisition of DeepMind for $400 million a decade ago served as a pivotal moment.
This level of activity attracted hype investors, eager to capitalize on the AI boom, pouring substantial funds into the field. This influx of capital created a chaotic landscape as investors raced to invest in AI companies. The hype cycle intensified, leading to restlessness among investors who were not directly responsible for their investments. As with every market trend, soon follows a tsunami of collateral damage, leaving a trail of failed ventures and lost investments.
Why do so many fail?
Many AI startups will inevitably become market roadkill simply because they fail to build out the fundamentals from the beginning. Eager to capitalize on the latest AI trends, they opt to front-end pre-existing platforms like GPT and Gemini. While this approach offers a quick route to market, it ultimately hinders long-term innovation and differentiation. These businesses completely bypass the critical stage of original thought and experimentation. While they may have impressive pitch decks and talented individuals, they lack the true lifeblood of any successful tech startup. Innovation should be ingrained in the DNA of the organization, driving every aspect of its operations.
The core of any successful AI startup is innovation. Bringing something new to the market is essential, as merely replicating existing solutions offers no competitive advantage. To achieve this, companies must hire individuals who are naturally curious problem-solvers and possess a unique perspective on the world.
Where innovation is your currency as a business, technology is your defense mechanism.
Is your business default dead or default alive?
To build a sustainable AI business, profitability and margin are essential. Excessive funding from VCs can actually hinder these goals, as startups become preoccupied with burning cash rather than focusing on profitability. By placing their future in the hands of VCs, startups risk becoming vulnerable to their investment decisions.
VC firms, acutely aware of the cash from their LPs burning a hole in their pocket, are under immense pressure to deploy capital quickly. This urgency often leads to hasty investments in early-stage startups, meaning this same pressure is passed onto entrepreneurs to deliver fast results. In the rush to launch AI products, many engineering teams resort to shortcuts, leveraging pre-existing models like GPT as a foundation rather than investing in original research and development. This approach, while expedient, leads to mass creation of generic, undifferentiated products. As the market matures, consumers will increasingly gravitate towards innovative, unique offerings, leaving behind the AI roadkill.
Ultimately, VCs view companies as assets with the sole purpose of generating substantial returns; numerous conveyor belts heading to the same destination. If a startup’s growth potential falters – and the conveyor belt slows down – VCs may quickly lose interest and invest elsewhere. The constant pressure to achieve milestones and secure additional funding can lead to unsustainable business practices.
Paul Graham of Y Combinator famously describes companies as being in one of two states: default dead or default alive. VC-funded companies are often in the default dead state, as they’re not making money.
Emerging from the hype cycle as a success
Once we filter out the roadkill, the AI market will undergo a major transformation. What will remain are AI technologies that provide genuine value and efficiency gains. Too many well marketed “AI” companies are creating solutions to problems that do not exist. AI will streamline processes, making it easier for businesses to scale operations. Contrary to the sci-fi portrayal of AI as a human replacement, its true value lies in augmentation.
By providing workforces with a robust exoskeleton, AI can empower businesses to continue striving for achievements currently deemed beyond their reach. However, it’s those AI firms that take the time to build out their own IP and proprietary technology that will stand the test of time and deliver the solutions that businesses come to rely on.
We’ve featured the best productivity tool.
This article was produced as part of TechRadarPro’s Expert Insights channel where we feature the best and brightest minds in the technology industry today. The views expressed here are those of the author and are not necessarily those of TechRadarPro or Future plc. If you are interested in contributing find out more here: https://www.techradar.com/news/submit-your-story-to-techradar-pro
