Your mailbox could explode thanks to Gmail’s generative AI

AI Basics

  • Google’s new generative AI composes emails.
  • It also summarizes long, pointless email threads, like those generated by AI.
  • If it works well, it can save you a lot of wasted time.

I intend to create the best artwork possible with my head, hands and heart. / Getty Images

Gmail’s new generative AI features bring every office worker’s nightmare to life.

A lot of office work is hectic work, and dealing with pointless emails is already a big waste of time. But with the new generative AI features in Gmail and Google Docs, you can now almost automate the act of your co-workers neglecting their work. So is this a nightmare or a productive life raft?

“Generative AI will undoubtedly have many unintended consequences, and increased email volume could be one of them.” Tom Tauri, author of the book “AI Basics,” told Lifewire in an email. “Ironically, generative AI could help address this to some extent. Summaries are a key feature, so doing this with incoming emails may become more common.”

Generative AI: Inbox Hero

New generative AI features in Google Docs and Gmail are currently being tested as part of Google Workspace. You give the AI ​​a text prompt and it composes an email out of it, creates a report, and so on. Also, it’s not just text. It can also create images for presentations, “allowing you to bring your creative vision to life with auto-generated images, audio, and video in Slides,” said Johanna Voolich Wright, vice president of products for Google Workspace in a blog post. said in

ChatGPT’s problem with this kind of email is that what it really struggles with is specificity and detail, and these are the most important parts of any worthwhile work email.

The possibilities are huge. All the heavy lifting of a cumbersome enterprise should be done by computers, not by humans. The dream is that instead of spending days writing a presentation about a project you’re coding, you can ask Workspace AI to draft a presentation and then proceed to code further.

But as we’ve seen, AI is much more than that when it comes to accuracy and truth. A conscientious person should thoroughly read, scrutinize, and edit AI-generated text for errors. But a frantic corporate executive might change the tune and send it on its way.

Talkative AI

As a journalist and former editor, I have had the occasional pleasure in editing less conspicuous copy. In such cases, it’s literally faster, easier, and better to ditch the original and write from scratch. So editing AI homework can be more work. And unlike humans who can provide feedback and guidelines for future articles, AI will be just as enigmatic every time.

And then we get to the point of this article, the email. AI is about to turn the rain of meaningless work into a tropical storm if you already have a co-worker to do the work for you. Here’s how easy it is to generate a few paragraphs of corporate terminology that are fully buzzword compliant. This still from Google’s introductory video shows what happens when you reply to an email and type “I’m on it.”

The original prompt “I’m on it” would have been much better.

“To date, most AI-generated text has been used for marketing and outreach tasks rather than for internal communication. is the hardest part of emailing any worthwhile job,” Michael & Associates attorney Ben Michael told Lifewire in an email.

However, there is a twist. One of AI’s capabilities is to summarize email threads. You can probably see where this is going. Instead of reading stacks of emails from various sources (AI and humans), you can simply tell AI to summarize threads.

What a beautiful irony if the bot were to summarize the flapdoodle above to “I’m on it.”

Busakorn Pongparnit/Getty Images

Legal Complexities of AI

There are other drawbacks as well. Like any tool, this can be used for both good and bad.

“Generative AI makes it easy for hackers to craft phishing emails, which is especially important for people with poor English,” says Taulli.

Also, although beyond the scope of this article, there are concerns about copyright, plagiarism, and other legal landmines when using AI to generate words and images. This is because AI is trained on vast amounts of existing human-generated data. -material.

What if AI plagiarizes existing text and disguises it as the company’s official business? Or generates a report with recommendations that turn out to be false or even dangerous? Are you responsible? I don’t know, but I would never ask an AI for an answer.

Source link

Leave a Reply

Your email address will not be published. Required fields are marked *