6 Startups Changing Everything with AI

Some technologies go far beyond being merely innovative, altering not simply their respective industries but the way society as a whole functions. Two great examples of this over the past few decades have been the arrival of social media platforms and smartphones, both of which fundamentally re-shaped how we live, work, date, and generally exist.

The sudden ubiquity of AI promises to be an even more seismic step for society.Some technologies go far beyond being merely innovative, altering not simply their respective industries but the way society as a whole functions. Two great examples of this over the past few decades have been the arrival of social media platforms and smartphones, both of which fundamentally re-shaped how we live, work, date, and generally exist.

The sudden ubiquity of AI promises to be an even more seismic step for society. The impact has already been so great that it’s almost surprising to remember that nobody had even heard of ChatGPT this time last year. Within two months of launching, the software was already a household name, racking up more than 100 million users.

Understandably, most attention has been given to the sensational creative possibilities of generative AI, whether that’s ChatGPT’s uncanny ability to emulate human dialogue and churn out instant articles on pretty much anything, or the way that eye-catching artworks can be conjured up using tools like Midjourney and DALL-E 2. 

But there are plenty of startups utilising AI in less “showy” but highly significant ways, and the disruptive possibilities are immense. Let’s run through some that have caught our attention.

Using AI to optimise commercial property insurance underwriting is never going to grab the limelight in the same way as, say, using AI to instantly create a Picasso-inspired cubist image of Darth Vader. But the success of Exeter-based startup Intelligent AI, which recently scooped up almost $2.5 million in funding, is a reminder that machine learning is fast becoming integral to all conceivable industries. 

The company’s platform uses predictive algorithms and digital twin technology to map out risks associated with physical locations. As the chair of on investment fund put it, the use of artificial intelligence to help insurers and brokers predict property risks is a “global market with significant opportunities for growth.” So we expect to see plenty more startups appear in this space.

London-based startup Hazy is the first company to take synthetic data to market. This rather resonant, sci-fi-like term refers to datasets that are generated by algorithms rather than real-world events. Such data can be used by businesses to test their systems, validate vendors and optimise innovations without using sensitive, personal information, thus avoiding any potential security breaches. 

By utilising Hazy’s data anonymisation and AI-powered data creation tools, companies can move data between stakeholders, and to external vendors, without undergoing onerous sign-off and governance processes. The firm has raised $9 million so far this year, with Microsoft among the big-name investors.

The advent of AI is a game-changer for healthtech, and one standout startup is Teton. Based in Copenhagen, its technology alleviates the immense pressures faced by nursing staff by making it far easier to monitor patients in clinical environments. Its system compromises smart cameras which scan wards, automatically documenting and analysing patient activities. 

As well as acting like the eyes and ears of nurses, thereby reducing their workload, the system makes use of intelligent analysis that’s attuned to the specific needs of patients. This reduced the risk of false alarms, with alerts only being provided when absolutely necessary. Investor interest in Teton reflects the real concerns over staffing shortages in hospitals around the world, with the company recently landing $5.3 million in backing.

Rare diseases may impact relatively few people on a disease-by-disease basis, but the combined impact of conditions designated as “rare” is immense. Indeed, it’s estimated that around 3.5 million people in the UK alone are living with rare diseases, and many will have been frustrated by the far slower diagnostic journey that’s part and parcel of such lesser-known maladies. 

London-based AI company Mendelian is tackling an issue which, as well as wrecking lives, also costs the NHS billions. Its flagship tool is called MendelScan, which automatically matches patient records with the published diagnostic criteria for hundreds of rare diseases. Reports are then created for healthcare providers, helping them diagnose far more rapidly than they otherwise might have. The simple but clever software has earnt Mendelian over $1.7 million in backing this year so far. 

Generative AI is synonymous with words and images (and, increasingly, videos). But machine learning can also be used to create biological materials, as the existence of Cradle.bio attests. The Dutch-Swiss startup, which received $5.5 million in seed funding last year, has created a platform which can – in the words of its co-founder – “substantially speed up the design, build and scaling phases when bioengineering proteins.”

The company is therefore ideally positioned for the big global push to leverage sustainable biological products, with programmable proteins now being used to create everything from lab-grown meat to petrochemical-free plastics. Experts such as analysts at McKinsey predict that over 60% of everything humans consume could be created using cell factories, so the possibilities for biotech AI startups like Cradle.bio are stratospheric. 

Adept’s sales pitch sounds as ambitious as anything we’ve heard. The company is “building a machine learning model that can interact with everything on your computer”, thereby functioning as an “AI teammate for everyone”.

In plain language, Adept is working on ways to empower software to not just respond to your commands, but actually carry out tasks for you. Its machine learning model, ACT-1, already knows how to use a web browser the way a human would. So, you could ask it to find you a house for sale in a particular area within a price bracket, and ACT-1 could use Google to do just that.

The applications go well beyond casual browsing. For example, ACT-1 could operate Salesforce on your behalf, carrying out logistical tasks that would usually require multiple clicks and typed commands in one fell swoop. This could represent the next big leap forwards for everyday AI: a kind of ChatGPT with the power to execute its own suggestions. Little wonder that the startup recently amassed $350 million in funding, hot on the heels of a $65 million fundraise last year.