Do startups have an advantage?

British start-ups and relatively small, innovative companies may have an advantage over larger companies in taking advantage of the generative artificial intelligence revolution, according to a report published last week by industry body TechUK and financial services company Experian.

The study predicts that generative AI could add as much as £120 billion a year to the UK economy, but only if companies successfully address a range of challenges including skills shortages, employee resistance and poor data quality or incompleteness. . As a side note, the study found that young technology-oriented companies are likely to adopt faster than their longer-established counterparts.

A good example is SteetBees. Founded in 2015, the company collects and analyzes customer behavior data on behalf of its clients. CEO Vidisha Gaglani says the company conducts what she calls “conversational research.” Customers are asked open-ended questions about their experiences with brands, and they can add videos and photos to their answers. The goal is to gain deep insight into consumer behavior.

Streetbees launched the platform well before the current wave of generative AI technology, but Gaglani says it has caught on quickly. “Over the last year, the platform has evolved and we use large language models (LLM),” he says. “We use LLM to almost automate complex research work.”

In practice, survey generation is automated, as is AI writing reports and conducting interviews, albeit under human supervision. “We are also able to generate executive summaries,” says Gaglani.

Gaglani says the combination of AI automation and human involvement enabled Steetbees to work on surveys faster while maintaining the depth and quality of the data. This is an example of how generative AI is already changing the way companies operate.

What about the bigger picture? Marketing companies were among the first to implement generative AI – often to create personalized marketing messages on a scale that would otherwise not be possible. But is the Experian/UKTech report right to conclude that innovative small businesses are better equipped to leverage technology than large corporations?

Everyone can play

Nick Hall is CEO of Experian’s GenAI Center of Expertise. He says companies of all sizes are aware of the need to move quickly, initially experimenting with ways to incorporate AI technology into their operations. “In this context, it is always said that startups are better at moving quickly than large corporations with hundreds of thousands of employees.”

Equally important, multilingual models should be used. “It’s a field that has been largely democratized,” he says. A small, two- or three-person start-up has access to the same LLMs as, for example, IBM or Experian, he adds. “They can buy the same enterprise product and have access to the same level of intelligence.”

However, he believes there are areas where start-ups can pose a challenge – skills being an example. “Tech start-ups may have the skills they need. But that’s not necessarily the case if you’re a startup operating in retail, for example,” he says.

Good data is also important. “All this technology must be powered by high-quality data. Even if you run a small business and just want to use generative AI to create product copy or design a marketing campaign, it will do much better if you equip it with high-quality sales and customer information, says Hala.

Buying or sellingG

Some distinctions need to be made. Some startups will create AI-based tools for sales. Others will be the buyers of these tools. “A lot of money is being made right now by companies creating AI tools,” Hall says.

Hence the high share prices of large technology companies and the almost feverish interest of investors in start-up companies that are at the forefront of data analytics in general and artificial intelligence in particular.

Tom Henriksson is a general partner at early-stage VC OpenOcean. He believes there is still a lot of caution in the VC community, perhaps with the exception of artificial intelligence.

Investments in AI start-ups

“There is currently a lot of capital in VC funds. A lot of large funds were raised, but they were not used very aggressively. There’s plenty of money to use, but you still need to be careful. However, artificial intelligence is very popular nowadays. It is seen as the next big platform after the cloud.”

So much so that even a round of seeds can look pretty like Stars. Henriksson cites the example of H, a French LLM developer that recently raised $220 million. “At the moment, the AI ​​infrastructure layer is still being built. We have OpenAI and Generative AI platforms that require huge investments,” he says.

But other categories are opening up. OpenOcean is particularly interested in startups that help large enterprises implement artificial intelligence. “These could be software tools, specific technologies or approaches. For example, helping enterprises implement AI that can be trusted and will not be biased. Helping engineers implement AI in their datasets,” he says.

Then there are vertical solutions. AI tools designed for specific industries, e.g. banking. Henriksson says this level of AI startups will require founders who not only understand data analytics but also the specifics of the sector. “True vertical solutions require a deep understanding of your market. So it may turn out that people who worked in banking will start cooperation with data analysts.” Henriksson says this is a corner of the market that is less developed.

We are probably only at the beginning of the generative AI revolution. At least for now, startups developing AI solutions and those putting the technology at the heart of their business models are well-positioned to attract investor attention. Meanwhile, small and technology-oriented companies can use tools that are already widely available. Those who are agile enough can gain an advantage over larger rivals.