In this issue

• AI cost 4,000 jobs last month

• How to use ChatGPT to plan a persuasive proposal

• Signs Microsoft may be planning price hike for AI

• Lawyer sanctioned for outsourcing research to bot

• Why we created The AI Writer

MAIN STORY

AI blamed for almost 4,000 job cuts in May

A major monthly survey of the US job market found 3,900 jobs fell to AI last month. It’s the first time researchers have seen AI cited as the cause of layoffs. But AI also creates opportunities for career progress.

AI caused almost 4,000 job cuts in the US alone last month, according to new research.

The study by employment consultants Challenger, Gray and Christmas reveals that employers laid off just over 80,000 people in May. Of those, 3,900 losses (5%) were because the roles were taken by AI automation.

A spokesperson for the firm told The AI Writer that this is the first time its monthly research has noted AI as the cause of redundancies. The report collates layoff data from company press releases, news media stories, official company filings and other sources.

In March, Goldman Sachs forecast that generative AI (such as ChatGPT) could replace one in four jobs and that two thirds of all roles could see some level of AI automation.

While ChatGPT grabs the headlines in the general media, AI has been quietly making big inroads into the workplace. As we reported last week, some companies are already using it extensively for customer communications. And law and accounting firms are investing heavily in the technology.

IBM also plans to pause hiring for roles that AI could handle. CEO Arvind Krishna said last month that the technology could replace around 7,900 jobs at the company. It plans to halt or slow recruitment in non-customer-facing roles such as HR, and around 30% of all IBM jobs could fall to AI in the next five years, he said.

The US-based National Eating Disorders Association also recently announced that it was firing all its helpline staff and replacing them with a chatbot. Lauren Smoller, a VP at the charity, blamed the change on a dramatic increase in helpline calls after the Covid-19 pandemic. (The announcement came shortly after the staff voted to unionise.)

New opportunities

But this is only half the story. Other commentators point out that any new technology tends to cause redundancies at first but creates new opportunities over time for the people and companies who embrace it.

‘We won’t be using less technology, that’s for sure,’ says Mark McDonald, a director at industry analysts Gartner. ‘We’ll be using more, different types of technologies. These are technologies that require new skills.’

Consulting firm McKinsey said on Tuesday that half its workforce of 30,000 people are now using generative AI, including ChatGPT.

A study by Stanford University and MIT Sloan School of Management of more than 5000 customer service agents found AI increased productivity by up to 14%. But it’s how it did it that’s interesting.

While the agents were having protracted live-chat sessions with frustrated customers, trying to work out what was wrong, the bots scanned the text of the conversations and suggested solutions in real time.

Human + AI for best results

The agents could then choose whether to adopt the solutions or not. In other words, it was humans working with AI that created the real win, not outsourcing either their thinking or their entire job to the new technology.

The biggest productivity increases were among the least-experienced agents. In the process, they also acquired new knowledge much more quickly than normal.

AI disruption, like any mass rollout of technology, is bound to be unsettling. The best way forward is to find ways to use it to accelerate progress, be that for your organisation, your career – or both.

As a young editor in the 1990s, I witnessed mass layoffs first-hand when a different new technology – desk-top publishing – wiped out the typesetting industry. But I also found myself eagerly snapped up by magazine publishers after becoming proficient in using page layout software.

One thing’s for sure, though. Saying, ‘I don’t really get technology. It’s not my thing’ could become a real career killer for all except those in manual roles. Because even if AI isn’t coming for your job, someone who knows how to use it is.

HOW TO

Use ChatGPT to plan a persuasive proposal

The classic, tried and tested Four Ps structure still works as well as it’s ever done. And now you can use ChatGPT to do the hard work for you.

Emphasis has been teaching the Four Ps structure in its proposal-writing training ever since I started the company 25 years ago. Tech may have changed since then, but humans haven’t. So it still works as well today. The difference now is that AI can help you build a persuasive argument in seconds using the same tried-and-tested method.

What to do: This works for ChatGPT or GPT-4. Just copy, paste and adapt this prompt:

I’d like you to produce a structure I can use for a proposal. The proposal is aimed at [job title]. Their main priorities are [priorities]. Follow the Four Ps structure: Position, Problem, Possibilities, Proposal. The Position section covers their current situation. The Problem section explains why that situation needs to change. The Possibilities section is where I will outline the possible solutions. And the Proposal section is where I will go into detail about my recommended solution.

I am going to give you a list of all the points I’d like to cover. These points are in a random order, and you do not need to follow it.

Take these points and generate a proposal outline for me based on the Four Ps structure.

Here is the random list of points:

[Copy and paste in your list]

Important: Be very careful about uploading private data into ChatGPT and be sure to follow your organisation’s guidelines. Even though the app now features a privacy option in its settings, it’s still wise to anonymise the information you type into it.

MICROSOFT 365 COPILOT

Microsoft pondering price hike for AI, claim industry insiders

The test companies in Microsoft’s trial of its Copilot AI are said to be paying up to 40% more for the privilege. The tech giant is reportedly weighing up whether to charge extra for AI or to raise prices overall.

Last week, we reported on Microsoft’s impressive plans to put AI in everything, everywhere, all at once. At the time, we noted that the company had yet to say what the new features – collectively called Copilot – would cost users.

It’s since been reported that many of the companies in the tech giant’s test group – which includes Walmart, Bank of America and Accenture – are paying at least 40% more than for standard Microsoft 365.

At least 100 of the organisations are stumping up a flat fee of $100k a year for a trial licence, which gives them AI access for up to 1,000 users. That’s on top of the $264k it costs for those people to use the standard software (although very large organisations may be paying less).

The claim was made in an exclusive article published by The Information, a well-respected Silicon Valley newsletter with deep connections to the tech world. It’s not clear if all those in the test group are paying the same price. But the piece quotes an inside source as saying that Microsoft is considering two price options: charging for AI features as an add-on, or including it for all users but raising the standard price.

The computer processors that providers need to run AI are expensive, and companies in the test group are said to be using the technology far more than Microsoft had predicted.

Zendesk, one of the most popular providers of customer-service software, is charging a $50 monthly premium for every user of its optional AI tools. This raises the cost by nearly 50% for some subscribers.

CAUTIONARY TALE

Lawyer in deep water after outsourcing research to ChatGPT

An experienced litigator used AI to help write a ten-page brief supporting a lawsuit against an airline. But unknown to him, it was a tissue of lies.

Last month, The Washington Post reported on how a driver followed her satnav straight down a slipway and into a harbour (still smiling, apparently).

But if you think that’s bad, AI can take ‘death by GPS’ syndrome to a whole new level, as the following story shows.

Not that GPS was involved this time. It was ChatGPT, which a New York lawyer with more than 30 years’ experience relied on a little too much for his research.

The lawyer, Steven Schwartz, was representing a man in a lawsuit against an airline called Avianca, which the man had alleged had injured his leg with an in-flight service trolley.

‘ChatGPT lied to me’ shock

According to The New York Times, Avianca urged the judge to throw out the case because it was too old and covered by the statute of limitations. But Mr Schwartz objected, citing case after case of relevant court decisions and quotations supporting the lawsuit, in a brief that ran to ten pages.

The trouble was, neither the judge nor Avianca’s lawyers could find the decisions or the quotations when they tried to look them up. There was a good reason for that: ChatGPT had made up all of them.

When the case collapsed, Mr Schwartz explained that he had never used ChatGPT before and wasn’t aware of its habit of being economical with the truth. He’d even asked the bot if the cases it had found were genuine.

‘Yes,’ lied the bot.

So, for the avoidance of doubt, never forget that ChatGPT is not human and has no conscience. It will not only lie to your face but do it so convincingly that you won’t suspect a thing.

Keep your brain engaged and use AI to help your thinking, not to do your thinking (or writing) for you. It can’t think, it just looks like it can. It’s like predictive text on steroids.

The lawyer was due to face a disciplinary hearing this week.

Why The AI Writer?

When I founded Emphasis, 25 years ago, it was just me, a cat and a kettle. It’s since grown to become the most trusted provider of business writing training in the world and helped more than 80,000 people, from 32 countries. We’ve worked with tech giants, top 10 law firms and major financial institutions, as well as at the highest levels of government. (We’ve even sent trainers to work with clients in the Himalayas.)

I would not claim that we’re experts in everything, as we’re certainly not. But, after working with the authors of around 100,000 documents, it’s safe to say that business writing is something we do know a thing or two about. And that includes witnessing all the various ways in which organisations get it wrong.

Now, we’re at the forefront of using AI to help them get it right.

Of course, I’ve seen nothing like the AI revolution that’s hitting us. Nobody has. But I do feel a responsibility to use our experience to help people navigate this brave new world of communication bots.

So here’s the deal. You get on with your job, while we immerse ourselves in the latest developments in workplace AI. We’ll worry about keeping up so you don’t have to. Then we’ll tell you exactly what you need to know to stay ahead of the curve.