Main Edition
US
FR
January 7, 2023 - 8:00am
In February, the Sun informs us, an artificial intelligence robot will break new ground in the legal world by defending a client in court. The defendant’s phone will relay proceedings to the AI’s server, and the defendant will say only what the AI tells him to, via an earpiece.
In the UK this would almost certainly amount to a contempt of court. So I was not surprised to discover, from the paywalled New Scientist article from which the Sun draws its report, that the project’s leader is keeping the court hearing’s location a secret — somewhere, apparently, “where this set-up can be classed as a hearing aid”.
The bold entrepreneur behind the exciting venture is 27-year-old Joshua Browder — son of Bill, the Anglo-American financier who has campaigned for multilateral sanctions against corrupt regimes ever since his tax adviser Sergei Magnitsky was beaten to death in a Russian jail. The plan originates from, and promotes, an app the younger Browder launched in 2015 while at Stanford University, called DoNotPay. It claims it can get you out of parking tickets and the like.
On the UK app store it has a rating of 3.1, with the word “scam” featuring heavily in the first several reviews. A 2021 puff piece records a valuation of some $200m, with investment from such luminaries as Andreesen Horowitz and, er, Sam Bankman-Fried. The latter does not currently feature on the investor list on the app’s Wikipedia page.
Having recently appealed a private parking ticket on the Independent Adjudicator Service website, I wouldn’t actually be surprised if the app could be of some use. The parking company’s submissions read very much like they were written by a bot, and they threw the towel in as soon as I responded to their final effort. Perhaps I could have written anything. But should you be taking legal advice from a chatbot at this point? No.
People often think they need a lawyer to put in all those credible-sounding “hereinafters” and other formal turns of phrase. Which I suppose they do, but that guff is really only there to beguile the laity. If you want to appeal a parking ticket you should just briefly explain why you shouldn’t pay and hope for the best.
Join the discussion
Join like minded readers that support our journalism by becoming a paid subscriber
To join the discussion in the comments, become a paid subscriber.
Join like minded readers that support our journalism, read unlimited articles and enjoy other subscriber-only benefits.
SubscribeOnce again, we’re back to Betteridge’s Law. To be fair, the author acknowledges this.
In my own professional area, I keep hearing that we’ll soon be replaced by robots. Been hearing this for a long time. Still have a job. Several systems that can apparently process complex, semantic information better than humans have come and gone; and they went because they were rubbish compared to a well-trained human. On the other hand, a good algorithm can cut through vast amounts of busy-work. I keep making the point that cutting out mindless, repetitive tasks in order to free up your expensive human staff is the way forward – AI should be your tool, rather than your master.
The Polish Army had a million horse mounted soldiers when Germany invaded – none had lost their jobs till than, and after – well, none still did that.
It is not a gradual process.
The situation of Polish cavalry is a flawed analogy. Horseback cavalry had been redundant since the invention of armoured cavalry and reinforced concrete defences more than 20 years previously. Poland had chosen not to modernise and was subsequently vulnerable. My experience of evaluating recent technological advancements and finding that (in some cases) the vendors promised far more than they delivered is a different experience. A better comparison would be that we were promised jet-packs and holidays on the Moon, but instead got Twitter.
The situation of Polish cavalry is a flawed analogy. Horseback cavalry had been redundant since the invention of armoured cavalry and reinforced concrete defences more than 20 years previously. Poland had chosen not to modernise and was subsequently vulnerable. My experience of evaluating recent technological advancements and finding that (in some cases) the vendors promised far more than they delivered is a different experience. A better comparison would be that we were promised jet-packs and holidays on the Moon, but instead got Twitter.
The people who work on this stuff are acting like the goal of human life is to eliminate humans.
I think you’re more or less right about that.
I think you’re more or less right about that.
The Polish Army had a million horse mounted soldiers when Germany invaded – none had lost their jobs till than, and after – well, none still did that.
It is not a gradual process.
The people who work on this stuff are acting like the goal of human life is to eliminate humans.
Once again, we’re back to Betteridge’s Law. To be fair, the author acknowledges this.
In my own professional area, I keep hearing that we’ll soon be replaced by robots. Been hearing this for a long time. Still have a job. Several systems that can apparently process complex, semantic information better than humans have come and gone; and they went because they were rubbish compared to a well-trained human. On the other hand, a good algorithm can cut through vast amounts of busy-work. I keep making the point that cutting out mindless, repetitive tasks in order to free up your expensive human staff is the way forward – AI should be your tool, rather than your master.
My first thought on this was around liability. A lawyer puts her/his name to a case or documents. There is a trail to liability when problems occur.
With a ChatBot where does liability lie? Insurance would be an interesting issue for the developer/owner/seller.
My first thought on this was around liability. A lawyer puts her/his name to a case or documents. There is a trail to liability when problems occur.
With a ChatBot where does liability lie? Insurance would be an interesting issue for the developer/owner/seller.
I got a chat bot spam phone call the other day. Sounded very human. They only way I could tell was that there were some odd ambient background noise shifts between responses and slight timing irregularities. The inflection was amazingly good. It was responding reasonably to my answers but the give away was it was also responding reasonably to my insults.
I need one of these AI to handle these calls for me. Now, there’s a product that would sell.
I got a chat bot spam phone call the other day. Sounded very human. They only way I could tell was that there were some odd ambient background noise shifts between responses and slight timing irregularities. The inflection was amazingly good. It was responding reasonably to my answers but the give away was it was also responding reasonably to my insults.
I need one of these AI to handle these calls for me. Now, there’s a product that would sell.
Good luck to the first tranche of lawyers willing to risk the ire of the regulators when this all ends in tears. People get struck off for putting money in the wrong bank account, let alone giving a chatbot control over a case.
Good luck to the first tranche of lawyers willing to risk the ire of the regulators when this all ends in tears. People get struck off for putting money in the wrong bank account, let alone giving a chatbot control over a case.
Leading to AI based judges enforcing AI produced law. Are we to be defended, prosecuted and judged by machines…? How would we appeal?
Well, if the bot is Android based then an Apple app will be used for the appeal and vice versa!
Well, if the bot is Android based then an Apple app will be used for the appeal and vice versa!
Leading to AI based judges enforcing AI produced law. Are we to be defended, prosecuted and judged by machines…? How would we appeal?
Not, I suspect, if lawyers have anything to do with it – they are the ultimate closed shop.
Yea, but there is no honour amongst them – so expect this to be the norm in 10 years.
Then your Starbucks Barista can hols a PHd ‘Women’s Studies’ degree, and the till operated by an ex-contract Law Solicitor.
£300,000 of student debt between them, so do not forget to tip……
They’re probably trying to find the paralegal version of this right now. They can’t find enough paralegals anymore.
Yea, but there is no honour amongst them – so expect this to be the norm in 10 years.
Then your Starbucks Barista can hols a PHd ‘Women’s Studies’ degree, and the till operated by an ex-contract Law Solicitor.
£300,000 of student debt between them, so do not forget to tip……
They’re probably trying to find the paralegal version of this right now. They can’t find enough paralegals anymore.
Not, I suspect, if lawyers have anything to do with it – they are the ultimate closed shop.
If your lawyer is Jolyon Maugham then you could replace him with a pedal-bin.
Which Harry could then converse with – he doesn’t need chatbots.
Which Harry could then converse with – he doesn’t need chatbots.
If your lawyer is Jolyon Maugham then you could replace him with a pedal-bin.
AI could soon replace the family doc. A machine can scan the entire internet for treatments and symptoms etc.
The problem with scanning the entire internet is that there is a lot of lies and misinformation on it. I think that AI will be an important tool to assist doctors in their diagnosis and treatment but not to replace them.
I think the point being made is the NHS (or private hospitals) could have a specialist-approved database which would be better than a GP with more patients than they can handle and we wouldn’t have to pay thousands of inflated salaries and pensions. It wouldn’t be checking “the whole internet”, just doing the job of a GP a billion times quicker.
I think the point being made is the NHS (or private hospitals) could have a specialist-approved database which would be better than a GP with more patients than they can handle and we wouldn’t have to pay thousands of inflated salaries and pensions. It wouldn’t be checking “the whole internet”, just doing the job of a GP a billion times quicker.
The problem with scanning the entire internet is that there is a lot of lies and misinformation on it. I think that AI will be an important tool to assist doctors in their diagnosis and treatment but not to replace them.
AI could soon replace the family doc. A machine can scan the entire internet for treatments and symptoms etc.
All the staff in the typing pool and telephone exchange assure me this will never happen.
All the staff in the typing pool and telephone exchange assure me this will never happen.
I enjoyed cross-examining ChatGPT on sex and race political questions and inevitably the answers came across as somewhat robotic and any difficult questions were evaded by generalised formula rather than properly addressed so I will not be in a hurry to entrust any legal cases to an AI lawyer until the programs are considerably more sophisticated than they are at present.
at the exponential growth in sophistication of these things – no hurry, you have 6 to 10 years yet, till they totally beat the human every time.
at the exponential growth in sophistication of these things – no hurry, you have 6 to 10 years yet, till they totally beat the human every time.
I enjoyed cross-examining ChatGPT on sex and race political questions and inevitably the answers came across as somewhat robotic and any difficult questions were evaded by generalised formula rather than properly addressed so I will not be in a hurry to entrust any legal cases to an AI lawyer until the programs are considerably more sophisticated than they are at present.
As a lifetime contrarian and rebellious by sheer instinct – what will my IA chatbot equivalent be like?
Or will they all basically agree?
It won’t matter because you (and I) won’t be around.
This is the most important question raised by all this AI nonsense. They will all agree and any contrary thought will fall by the wayside. So there will be no more Eurika! moments, no more flashes of intuition no more artistry. It’s hard to think of a human endevour, from brain surgery to plastering a wall that won’t suffer.
There’s a famous and beautiful bit of Hemingway, the first paragraph of “A Farewell to Arms”; in one short passage, just three or four sentences, he used the word ‘and’ more than a dozen times. Can anyone imagine an AI program that wouldn’t ‘correct’ that ?
It won’t matter because you (and I) won’t be around.
This is the most important question raised by all this AI nonsense. They will all agree and any contrary thought will fall by the wayside. So there will be no more Eurika! moments, no more flashes of intuition no more artistry. It’s hard to think of a human endevour, from brain surgery to plastering a wall that won’t suffer.
There’s a famous and beautiful bit of Hemingway, the first paragraph of “A Farewell to Arms”; in one short passage, just three or four sentences, he used the word ‘and’ more than a dozen times. Can anyone imagine an AI program that wouldn’t ‘correct’ that ?
As a lifetime contrarian and rebellious by sheer instinct – what will my IA chatbot equivalent be like?
Or will they all basically agree?
“an AI could pen a more persuasive jury speech”
It would be rather amusing to see a human lawyer and AI compete in such a Turing Test.
“an AI could pen a more persuasive jury speech”
It would be rather amusing to see a human lawyer and AI compete in such a Turing Test.
ahh yes.. slister Kevin Chatbott in his bri nylon short £18 Tesco suit and pointy corfam lace ups, replete with reound veowel seound estuary accent…. hang on a minute… he is a robot anyway…..
ahh yes.. slister Kevin Chatbott in his bri nylon short £18 Tesco suit and pointy corfam lace ups, replete with reound veowel seound estuary accent…. hang on a minute… he is a robot anyway…..
Chatbots are alarming, but I suggest here one way in which they could be helpful: writing diversity statements. Not only can they be handled in seconds, only accepted views will be expressed. Subversive? Yes.
https://www.youtube.com/watch?v=QJJEU-BfQR0&t=12s
erikhildinger.com
Very happy I’m not qualifying as a young lawyer right now.