Insights & Trends

What does defensibility look like in AI?

Insights & Trends

What does defensibility look like in AI?

Words Bianca Furtuna

September 6th 2023 / 8 min read

Is the AI hype cooling?

Compared to six months ago, when AI dominated the agenda and new businesses seemed to be launching, raising, and scaling over shorter time frames than ever before, we’ve seen the tide changing. ChatGPT, at one time the fastest platform to a million users, has started seeing a drop in user numbers, while Jasper AI, who raised a $125 million Series A last October, announced publicly they’d be laying off staff.  

The reality is that as the hype fades, stabilisation will start to set in. It’s in this period that we’ll start to see the best, long term use cases of AI and machine learning starting to emerge. Those businesses that fail to build a moat around themselves will no doubt suffer as a result.

But with a new era of technological innovation comes a fresh set of requirements for building a defendable business. Here’s what I think will help you see off the competition in a crowded AI market.

Interested in receiving more insights like this ? Subscribe to our newsletter and join 17.5k founders, investors & innovators.

Subscribe here

A proprietary dataset

“If a company has proprietary data on top of which it builds an AI programme to interpret the data, then the company’s AI-enabled proposition has defensible value.”Sulabh Soral, Chief AI Officer at Deloitte

Data is typically considered a moat for any business, something which has up to this point established the dominance of a small number of Big Tech firms. With regards to AI, data remains a huge differentiating factor—specifically, the dataset on top of which you build your foundational model, and your strategy for protecting that data. 

There are a number of key factors to consider within this. First of all, does your dataset give you unique insights that no one else has? This may be related to your unique scale or penetration into a certain industry, or it may be that the dataset you have is particularly difficult to gather. 

A ‘wrapper’ on top of a foundational model, such as ChatGPT, does not have this data defensibility since there is nothing stopping someone else building the same tool quickly and efficiently without having to invest into gathering data. These business must find defensibility elsewhere.

Built AI is one startup leveraging the benefit of a proprietary data model, having created a detailed digital twin of the world’s real estate market to help drive decisions in the commercial property industry. Similarly, Qumata offers a new standard for life and health insurance underwriting through its vast proprietary health data set that can accurately calculate risks for any individual and over 800 medical conditions. 

It’s also important to consider how the value of this dataset may evolve. On the individual level, it may be this improves over time for each user (for instance, with a model that is trained or tuned to the behaviour of each user). On a model level, it may grow as the number of users grows (a classic network effect). Reflect on what it is that you are building and whether you can build this defensibility into it. Not all businesses will have this opportunity: models that understand and predict behaviour can constantly fine tune, as behaviour changes over time; whereas speech recognition has a ceiling of how much it can improve.

Feedback loops (where data is fed back in and improves the model) using AI generated outputs, moreover, don’t really behave in the same way that they might with other data sets. On the contrary, some have raised alarms around the risks of training a model on model-generated content, leading to potential model collapse. 

Going vertical, not horizontal

Rather than going for a broad market, AI startups may be better placed specialising down a vertical and serving one particular domain. See how AutogenAI has built a generative AI tool for writing bids and proposals; or Xapien, who call themselves the ‘ChatGPT of due diligence’, condensing weeks of background research into minutes.  WiseWorks has brought AI to financial institutions, helping them extract insights and value from workplace communications via natural language processing. 

Integrating deeply into a particular vertical has great potential for defensibility. Larger organisations will always beat startups to the mass market, so you are much better trying to serve the long tail by pinning down a niche. Try to avoid developing something that a larger business can build as an extension of their existing product. 

This requires domain expertise (e.g. someone on your team with experience and/or network in that vertical), to ensure you understand well the processes and pain points of the user and where AI can make their job more efficient. 

The use case is important too. AI is particularly useful for certain types of tasks—repeatable, highly paid, ones which allow small room for imperfection (acknowledging the imperfect nature of existing models), and where digitised workflow does not yet exist. This last point is  a huge opportunity for defensibility, but can be hard to nail: crucially, you need to create ‘stickiness’, so make sure your product can easily become embedded in their workflow.

Build deep tech at the core of your IP

Another aspect of defensibility is the technology itself—is your product or IP of unique value that means it's doing something that no one else can do? 

There are a number of paths towards this. Building your own foundational model means that your tech has a key differentiation point that is hard to replicate. This is particularly true if you are developing new models with a difference (e.g. filling gaps from other models). Mistral, for instance, has an open source lens to its model; Poolside, meanwhile, aims to build a tool that can write software using natural language. Along the lines of going vertical, not horizontal, founders may explore opportunities to build foundational models for different domains. 

Another route to defensibility is building an AI system that automates the discovery of solutions in complex deep tech spaces, such as biotech or material sciences. Take Peptone, a startup using AI to discover new drugs for complex disordered proteins that typically might take decades to discover. Or there’s Materials Nexus, a startup accelerating the R&D process for ‘next-gen’ materials using AI and quantum computing. 

If you are building in a complex, deeptech space, your team brings enormous value to your proposition. Having domain expertise, particularly in a scientific space, adds weight behind your business, which will help enormously when it comes not just to your engineering process but also to sales, fundraising, and so on. 

Step up from demo to product

Did you ever download Lensa—the AI photo app that allowed you to generate dozens of stylishly edited portraits for a small fee. A solid hook and strong marketing campaign meant millions downloaded the app in its first few months since launching. 

Lensa’s failure, however, was in never building on top of this ‘demo’. A two-dimensional feature that got people’s attention in the first place failed to keep people engaged, leading to a sharp dropoff in users. 

Building defensibility requires consideration for how to elevate your demo into a product. This is particularly important if your AI is just one component of your product (e.g. sits alongside other features). 

At its core this is about identifying a really strong problem, and building an AI system which automates the solution. You shouldn’t just be building AI for AI’s sake. On top of this, you need to ensure you have a really strong GTM strategy and scalable business model.

Look beyond generative AI

Amid the hype around generative models, it may sound glaringly obvious but founders may be better served taking the path less travelled. 

Consider what it is that generative AI does, and more importantly, what it can’t do. There are still enormous problems to be solved, and a number of strong machine learning techniques that can help solve these.

Broadly, there are two types of model—generative and discriminative. Generative models take huge swathes of data and information and apply them to generate new data points: for instance, large language models that are capable of producing text (ChatGPT) or diffusion models used to generate images (Stable Diffusion). 

Discriminative models, meanwhile, are able to distinguish between the different data points and draw boundaries between them using probability estimates and maximum likelihood. There are numerous opportunities here that generative AI cannot touch—including regression models, classification, forecasting, recommendations, computer vision, and so on—where building a proprietary ML model can set you apart from the competition. 

Over the next few months, as some of the hype around Gen AI dies, expect to see the next wave of AI companies coming along—those applying innovative technology to interesting problems will be best positioned.

Some final considerations

There is a counter view, potentially contradicting what I’ve written up to this point—that in fact, most early stage startups are non-defensible. Eminent tech investor Elad Gil writes that in these early stages, founders should be more focused on serving their customers’ needs exceptionally well. From this point, it's the decisions you take next that will determine if you build defensibility: essentially, how well you build, execute, expand, ship your product, take on feedback, develop, build new products, etc. 

So while founders may be predisposed to thinking about defensibility, it may actually be worth considering that defensibility builds over time. Building in key areas can help differentiate you and set you apart from the hype, but the fundamentals matched with good execution will ultimately be what builds a moat around your AI.

Building a business using AI?

Find out more and apply to one of our Accelerator programmes

Apply here

About Bianca

Bianca Furtuna is Head of Data Science at Founders Factory. Bianca spent several years at Microsoft leading a team of Applied ML Scientists, delivering complex AI solutions using Responsible AI principles. With experience across several industries & spanning many data science areas, she is passionate about making the transformational potential of AI a reality.

Words by
Share article

News from the Factory Floor

factory news

Investing in Zenus AI—ethical facial analysis for events & retail

We're excited to announce our latest accelerator investment with Pico, and how they deliver deeper behavioural insights for the event and retail industries

factory news

Investing in Connexa—digital & telematics-first motor insurance

Learn more about our second MB Speedup investment, and how they transforming digital insurance for traditional providers

factory news

Investing in Dataspark—the GenAI ecosystem for enterprise

Learn more about our first MB Speedup investment, and how they’re helping large enterprises (particularly in finance) tap into the power of generative AI