The rapid growth and investment in generative AI have sparked a race among startups to leverage this technology and build AI-native applications. In Q1 of 2023 alone, US$1.7 billion was invested in 46 generative AI companies. However, in the midst of this excitement, it is crucial for these startups to consider the long-term viability and competitiveness of their products.

As we reflect on the transition from the introduction of horizontal cloud software in the early-2000s to the emergence of industry-specific cloud software in the late-2000s, it becomes clear that specialised startups hold immense value.

Building defensibility

We see a few key considerations to think deeply about as startups think about how to build defensibility vis-a-vis the models and potential incumbents coming in. To avoid getting commoditised, there are two critical questions: does the startup provide enough value on top of the model layer? And on what basis does a startup develop its moat?

One of the fundamental considerations for startups is to assess whether they offer enough value on top of the model layer to avoid being commoditised by it. While generative AI models provide a strong foundation, startups must focus on incorporating AI value through prompt engineering and fine-tuning specific to their use cases.

However, it is equally important to evaluate whether the value provided by the startup will remain significant as models continue to improve. Startups should strive to go beyond the capabilities of the underlying models and offer unique features, insights, or services that differentiate them from competitors.

Incorporating private data and customer context into the generative AI model is a powerful way for startups to improve the quality and relevance of their outputs. By leveraging proprietary data sources and understanding customer needs deeply, startups can enhance the accuracy, personalisation, and value of their AI solutions.

Also Read: Solarad.ai can forecast accurate energy generation for solar plants, battery storage

AI’s value will be most powerful when tightly focused, so startups should aim to accumulate proprietary data. This strategy enables startups to create a competitive advantage that is difficult for incumbents or generic AI models to replicate.

The age of generative AI

In the current landscape, generative AI applications are following a trajectory akin to early cloud companies, albeit with a crucial distinction. Unlike the cloud era, where startups like Salesforce, Workday, and ServiceNow emerged as ground-up pioneers, the underlying technology of generative AI now closely aligns with industry giants such as Microsoft, Google, and Meta.

Moreover, with the advent of APIs and open-source models, the adoption of generative AI technology has become markedly more accessible, fostering an environment conducive to both incumbents and startups. Consequently, as these major players dominate the broad horizontal applications, startups must recalibrate their focus toward specific domains with narrow contexts.

To maximise the value of generative AI, startups should consider the optimal insertion point for their product within existing workflows. This requires understanding the pain points, inefficiencies, or opportunities where AI can make the most significant impact. By going deep into specific workflows, startups can minimise disruptions while unlocking the full potential of AI.

In the age of generative AI, the stronger your context window, the wiser your model and product. Founders need to have a deliberate emphasis on the emerging technical capabilities of generative AI, accompanied by a relentless pursuit of identifying functions or vertical problems that stand to benefit from their unique insights.

We’re at an inflexion point. Startups must learn from past platform shifts and understand the importance of narrowing their focus and leveraging proprietary data to create defensible businesses.

By providing value beyond the model layer, incorporating private data, going beyond incumbents, and finding the right insertion point, startups can build specialised AI applications that deliver superior outcomes for customers. In the era of generative AI, specialised software holds the key to success.

—

Editor’s note: e27 aims to foster thought leadership by publishing views from the community. Share your opinion by submitting an article, video, podcast, or infographic

Join our e27 Telegram group, FB community, or like the e27 Facebook page

Image credit: Canva Pro

The post Unleashing the power of specialised AI startups in the era of generative AI appeared first on e27.