Will AI Tools Make Employees More Productive?

Say hello to ChatGPT Enterprise and Google Duet AI.

Johannes Sundlo

Sep 1, 2023

Welcome to FullStack HR, and an extra welcome to the 154 people who have signed up since last week.
If you haven’t yet subscribed, join the 5000+ smart, curious HR folks by subscribing here:

Happy Friday, folks,

Thanks to everyone who commented or reached out to me regarding the YouTube video I posted last week - it truly means a lot. I usually record these videos in late-night sessions, and it’s always appreciated when people reach out with thoughts and feedback.

This week’s video looks into whether it’s possible to use generative AI tools as salary benchmarking tools - will it?

If you watched my webinar or heard me speak, you’ve heard me speak about that one thing that is likely to happen is that we get proprietary versions of ChatGPT. Basically, it's a safe version of ChatGPT that doesn’t train their model, keeps your data safe, and keeps your CIO semi-happy.

And on Wednesday, OpenAI released ChatGPT Enterprise.
Their blog post alludes to compelling stuff for CIOs and security officers, such as robust privacy protections. The system is SOC 2 compliant with encrypted data and also offers new admin controls to enable large-scale rollouts.

Beyond that, there are some interesting data points as well.

For example, the adoption of ChatGPT has been swift, with over 80% of Fortune 500 companies using it. How did they come up with 80%? They looked at company emails associated with ChatGPT-accounts. (Someone hasn’t been watching my webinar…)
The main use cases OpenAI sees are crafting communications, accelerating coding, exploring solutions to complex questions, and assisting creative work.

It’s quite remarkable if you think about it. Here's a tool that is less than a year old and already used by most of the world's largest companies. If someone were to offer that adoption rate to any other SaaS solution, I’m sure everyone would say yes.

The launch of the enterprise version also combats one of the most common objections towards ChatGPT - security. That doesn’t mean I think we will see immediate widespread adoption and use of ChatGPT Enterprise everywhere. But there’s now less of a hurdle to do so, which might reinforce last week’s article - that those who dare to try and explore AI will gain competitive advantages.

And if you find that you need more data than a bald HR dude saying that there might be benefits to exploring AI, then I have good news for you.

BCG dropped this one last week, saying that HR productivity could increase by ~20-30% thanks to AI automation of routine tasks. A report from IBM showcases that companies evolving their operating models are way ahead. They already see 44% higher revenue growth and 33% better employee retention.

So, we can assume that there are real, tangible benefits of starting to explore how to work with AI.

Still, as said, few are digging into AI. And once again, this is not only the bald HR dude talking. This is from Gallup.

Only one in 10 employees report using AI at least weekly, while two in 10 use it less frequently. Seven in 10 say they never use it.

Most don’t feel prepared for AI. More than half of employees (53%) say they don’t feel prepared to work with AI, robotics, or other advanced technologies. Twenty-six percent say they are “not at all prepared.”
Most remain complacent about replacement. Even as tools like ChatGPT have accelerated AI conversations in the boardroom, few workers -- just 14% this year -- think it’s “somewhat” or “very” likely their job could be eliminated in the next five years due to new technology, automation, robots, or AI. That’s a little different from the 13% who said the same in 2019.
Most are also skeptical of AI’s potential. Most workers are skeptical that AI can be used to improve how work gets done. Just three in 10 agree it can be beneficial.

If this holds true, we do have some work to do, ChatGPT Enterprise or not.
But perhaps Duet AI will change that?

Google & Duet AI

The release of ChatGPT Enterprise on Wednesday was not a coincidence.
Later that day, Google hosted its Google Cloud Next conference, where they talked a lot about generative AI. And when I say a lot, I mean a lot.

AI was mentioned 262 times in the keynote, or 2.787 times per minute.

Almost everything revolves around AI, one way or another, and it’s clear that Google is investing heavily in this space.
Not only with Duet AI, which we’ll come to in a second but also with their Cloud offering as a whole. I know that you should always take company presentations like this with a pinch of salt, but it’s fascinating to see how AI is starting to power almost every single job, one way or the other.

One would be a fool not to see that this will impact our workforce.

Not saying that the impact will be bad or overstating the effect it will have; it might be like the digitalization prior to this, but that it will have an impact.

Because Duet AI is now more or less omnipresent in Google’s products, and if you are a Google Enterprise customer, Duet AI is now available to purchase for you. The price? 30 bucks per user.
Is it cheap or not? My guess is that the biggest blocker for widespread AI adoption now will be the CFOs…

Joking aside, if productivity increases in line with the estimates above (20-30%), then 30 dollars is a steal. But then you need to make sure you really increase productivity as well.

And you’ll likely will.

Why? Because Duet AI has been integrated into popular Workspace apps like Gmail, Docs, Sheets, Slides, and Meet. According to Google, it acts as a digital collaborator to help users with writing, visual design, meetings, and more.

Some real-world use cases highlighted during the keynote were:

Generate emails with the right tone, like reminders for late invoice payments.
Help write content like blog posts on complex topics like regulations.
Shorten the creation of sales and marketing documents by automatically pulling in relevant data.
Attend meetings and provide real-time transcriptions and translations.
Summarize meetings and provide catch-up help for those who joined late. (This feature seems great, but will this also mean people will be late as they can easily catch up?)

Going forward, Duet AI will offer more proactive, contextual help based on what users are currently working on. It may even take actions like scheduling meetings automatically.

And this is all done, at least if you listen to Google, in a safe and sound way - you don’t train their models, and they maintain the safety standards they already uphold. (After all, we’re a bunch of companies already using Google Workspace, 10m+ or so…)

What do we do now?

You’ve heard me saying this before. But I’ll say it again.

We need to be curious and educate ourselves.
(Hint: It will not be the last time I say this.)

I’m not sure if ChatGPT Enterprise nor Google Duet AI (or Microsoft Co-pilot) will solve all our problems, but they will be stepping stones. Dare to try, share what you are doing, and explore use cases for your organizations with these tools.

I look forward to reading and listening to all the stuff that’s happening out there!