Waymo shows he Legal Way – Artificial Lawyer

Waymo took over 20 years to finally give safe, accurate, self-driving vehicles. In San Francisco they are now 100% for real and artificial lawyers traveled to one. The legal one can learn a lot from this trip. Here’s how.

Tell us the way, Waymo

Waymo could trace his roots in about 2005 when a team of Stanford University researchers focused on the idea of ​​an autonomous vehicle gathered in California. In 2016 the group officially became Waymo, a Google result and the alphabet.

At the same time, again in 2016 the artificial lawyer began as a blog about the arrival of ‘he legal’, based on the belief that this technology would turn the legal sector.

Both Waymo (and other companies), as well as the pioneers of the legal one, have then spent many years and many money trying to make their original promises.

For Waymo the promise was to provide a safe and reliable autonomous vehicle that could sail even on the complex and sometimes busy roads of a city like San Francisco – completely on its own.

For the legal one …. To reach a point where he is able to absorb the course of all legal workwhere a human lawyer is largely a quality controller at the end of the process. Of course, this is for the most routine tasks, although processes that can be quite complex and include many non -structured dimensions and data. We still need lawyers to treat a wide range of anthropocentric needs – and are likely to always.

In short, both problems are very difficult. But, as Al, Al, Waymo can solve her challenge and did so with Aplomb. Jaguar – with which Waymo works – drives safely, convincingly and accurately. I felt completely safe.

Moreover, in about 10 seconds it already felt completely natural to travel this way. Perhaps unconsciously this site can feel the purpose of each maneuver of the autonomous vehicle while weaving smoothly through traffic in a very American way of using the road.

By the time it was pulled out of my hotel- (which was accidentally reached by a key card and then ‘operated’ through an iPad in the room) -I was 100% sold in self-driving cars, just as I was 100% sold on the iPhone within one hour of my first phone ever awake many years ago. It just worked. He felt intuitive. He felt reliable.

Now, we cannot always say the same legal one. For many cases of use, you really get that immediate sense of success and accuracy – which is one of the reasons why Genai is so quickly taken by so many lawyers. But ‘autonomous legal work’ is a difficult problem, and also extends in many ways, sometimes with exponential growth of difficulty.

For example:

• Summarize this short customer email – easy.
• Write a quick thank you to the client – easy.
• Now look at the legal question in that email and find the materials related to the issues – first from general legal resources – average difficulties.
• Now do the same, but from the data of this legal firm from the last ten years, but avoid files that should not be viewed by those not allowed. Also, do not attract everything. Show only two or three main issues of the past that really match the ‘ideas’ within the client’s email – becoming more difficult now.
• Now draft a convenient legal document that includes legal needs – at this stage of the process – of the client, making sure you use the right data, use ‘What’s Market’, stay within your firm’s way to work for such issues, and do not lose any major clauses, dates, names and other data, and make sure that every word is significant and so that it is significant. Correct – very accurate.
• Now, without seeking help for any human lawyer, engage with the client, get their reactions to the document and make your judgment on what should happen next and make any new changes. All the time, remember that this should be close to perfect accuracy and not contain errors or omissions that would affect the end result – very, very difficult.

But is it driving into the hyper-complex world of city traffic, with other cars, motorcycles, pedestrians, changing road appearances, pits, sometimes wet weather, mixed with coincidence and inevitable errors of human drivers nearby, and driving among other autonomous vehicles, more complex?

They are different types of the very difficult problem: one in the physical, structured environment of the 3D world, the other in the non -structured and subjective world of human understanding, viz. ‘Words on a screen’.

Waymo has solved his problem, though with 20 years of work and billions after billions of dollars spent on his solution. The legal one has not had so much time, nor did it have the same level of investment.

Al Ask Gemini (part of Google) how much it was spent on Waymo development. The answer was: ‘Some reports have estimated the total alphabet expenses per Waymo (including external funds) to be around $ 30 billion by July 2024. This suggests that a significant portion of the elevated capital has been constantly invested in the research, development and scaling of autonomous Waymo management technology. ‘

Then if you add investments made by many other companies in the same field, e.g. Uber, Tesla, and others, as well as investments in some of the basic technologies that supports these cars, from the latest to ‘physical’ backbone, then we have a giant investment globally.

But it is still less than what it cost, (in 2023 dollars), to set people in the moon in 1969, which was entirely about $ 300 billion.

How much has been spent on the development of legal AI? It is difficult to show, plus money is not always a perfect measure of potential success – though it certainly helps. Moreover, Genai’s ‘backbone’ today is not made by legal technology companies, LLMs are from Openai, Google, Anthropic and many others. In the pigs on them. So our investment is actually much smaller.

Neither can a legal technology company really control what Openai, or other model manufacturers, even if one or two people there have Sam Altman at fast speed. They can refine the results, they can probably be combined with some much smaller models and other large models, and they can be really good at Rag and ‘showing’ LLM included in a highly curated data collection… .but when it is for the most complicated use cases in law, we are still raising the state of art to get to where we need to get to where we should be.

That is to say, extending if we aim for ‘autonomous legal production’ – compared to it simply acting as a small friendly assistant that takes us only a small part of the road along a task and even then we do not fully accept it ourselves.

What does this tell us?

Personally, I found the Waymo experience a real catchment moment. Like everyone else, I had seen with interest, then disappointment, over the years after the promise after the promise of cars driving and then failed to surrender.

But now, in California, this promise has become true. And it was empirically providentally. I sat in a way out and took me home, safely and safely. It worked, for real.

And this gives the artificial advocate an increase in optimism where the right is going. While many lawyers are satisfied with LLM and related applications that provide a semi -road results that help, nor can they do a whole task, nor are they completely reliable, the goal should be on a Waymo scale, at a level of Waymo ambitions: enable a whole, very complex, legal and accurate flow.

I’m sure we will get there.

The only question is: How long will this last? But we will get there.

After that, the next question is: What impact will this have?

The PS is worth noting that some estimates suggest that Waymo’s automated taxis already make up about a quarter of all trips to San Francisco, and they have only a fleet of about 300 vehicles there currently-even though they work much longer than man-led taxis.

–

Many thanks Todd Smithline for the first journey of AL to Waymo. You can find more information about the company here.