Simple legal automation & the AI bubble

Here is today's good news: the next financial crisis seems just around the corner. Nouriel Roubini, who earned the nickname Mr Doom when he foresaw the last one, predicts it to come in 2020 and the latest update from Intensity estimates that there is 99.9% probability that a recession will happen within the next two years. With an ongoing US-China trade war and a British exit from the European Union in the pipeline, it is hard to Legal Tech Weekly to do anything about that. However, a global recession is not the only thing to fear. ‍

The AI bubble

‍Recently, there has also been much talk about an AI bubble caused by a discrepancy between the enormous expectations to the technology and it's actual capabilities. The expectations manifest in rapidly increasing investments and rising valuations that look similar to those in the dot-com bubble that famously burst in 2000. At the same time, 40% of Europe’s artificial intelligence start-ups even found to lack the AI they were pretending to have.

We are experiencing a rise of Pseudo-AI, where tech-firms hire cheap labour to do a robot's work. In average, companies branded with AI have a higher valuation and founding rounds, but actual AI is expensive to develop. So the companies simply “fake it until they make it, telling investors and users they have developed a scalable AI technology while secretly relying on human intelligence.”

These so-called Mechanical Turks are also present in the legal industry where AI is the buzzword of all buzzwords. (Here is a hint: If the intelligence only works during normal hours, it's probably not artificial). Knowing that the legal industry is often risk-averse and sceptic towards new technologies, that might actually end up slowing the adoption rate of legal technology and thereby causing harm to the industry as a whole. Luckily, some companies do it differently. Here is one of them...

‍Simply legal automation

‍“If a business spends a lot of recourses, money and time on a digital product which for whatever reason doesn't work or doesn't deliver promised benefits, that is likely to have a negative impact on the attitude to legal technology in that business ,” says Max Cole. He is one of three founders of the UK legal tech company Autto. According to him, simple automations can be even more beneficial than advanced artificial intelligence, especially in to small- and midsize law firms.

“The shiny, revolutionary, exciting aspects of AI and blockchain steal the headlines from the more mundane digital transformation tools that are available. I can see how machine learning and deep learning can be used by big law firms who have got the budget to train it on datasets and where the efficiency gains are obvious. But for most legal practices operating on a lower level, it is difficult to see right now what their ey AI use cases might be. That is not entirely true because I know there are contract management tools that successfully incorporate machine learning, and I can see how they can work and be useful. But for a lot of day to day tasks it is not obvious how AI is going to help at the moment,” Cole elaborates.

However, improving efficiency and productivity with automation does not require artificial intelligence. Autto's software platform uses simple automation to help businesses improve their processes where it is not economical to build heavy-duty RPA API-solutions or train artificial intelligences. With their tool, law firms and legal counsels can build theirre own automated processes with configurable building blocks that can be stacked upon each other to create unique workflows.

Since he has worked as a lawyer for longer than he “cares to remember” and also functions Barrister, he knows that automation where automation can be beneficial.

“Every lawyer will tell you that they do stuff in the course of their day to provide legal services to their clients which is routine. So let me give you a really basic example all lawyers will know: I frequently always have to produce this short document letter, and I then have to send it to these 5 people who when have to approve it and sign it off. I always have to do it the same way that and it requires no intellectual effort. I do it by finding what I have done before, then I copy/paste it and then have to go and find their email addresses and collect their replies. It's tedious. It's boring. It requires no intellectual effort but I have to do it. I know this because I have done it myself,” explains Cole.

While Autto could have tried to brand their expert system as artificial intelligence, they have chosen the honest approach to business: “We are proud of what we do. We think that simple automation is a valuable tool for legal services. We don't need to pretend to be something in order to be valuable to people. And once you start talking to people who understand AI and know what it is, they will say: “You are not really an AI-business”. If you go back into history, it would be classed as AI, but what AI it really means today is machine learning or deep learning. So it would be unauthentic and counter-productive to brand ourselves as something we are not,” Max Cole concludes.

In order to build a sustainable legal tech industry, we might need more honest companies like Autto.