How Google’s Gemini AI app is making me more human

A few months ago, I randomly reported to my editor that some chats on the Teams app were being powered by AI. I wasn’t using AI tools to fully automate all aspects of my work. Of course it is unethical and a blatant abuse of the contract.

Instead, I used Apple Intelligence to correct my writing, fix typos, and temper my em-dash enthusiasm. In two parts, I narrated the message, transcribed the audio with an AI tool, and proofread the wall of text using OpenAI’s GPT-4 wizardry.

It would have taken at least twice or even three times as long to type the same message on the phone. If my phone didn’t work, I would have had to stop on the platform, sit on the cold metal seats of the noisy station, and type my responses into my laptop.

As of today, AI is an integral part of my inbox and feels more “human” in some ways. Before I said yes to Gemini becoming a part of my inbox, I was heavily debating whether AI should exist in a space as intimate as Gmail.

Gemini is there. All it takes is a leap of faith

“Tell Tyler this idea is great, but I’m not covering breakthroughs in drug discovery. But tell him to keep sending pitches. I wish him luck. .”

This was my recent request to a public relations officer at a reputable scientific institution. He sent me a press release about a drug that started working on migraines almost instantly.

The email Gemini wrote was an elaborate reply conveying my excitement over the discovery while expressing my inability to write about subjects outside of my field of expertise or the publication’s coverage. I felt better after hitting the send button.

Until I used Gemini’s Compose feature in Gmail, I rarely wrote emails rejecting suggestions, even though I wanted to respond politely and maintain long-term contact with the sender. There wasn’t.

Normally I try to keep my responses as casual as possible, but if I change my mind, Gemini provides a route to formalizing words with just one tap. The quick actions “Elaborate” and “Shorten” are also useful.

I often include lines like “Try to keep your tone lighthearted” to elicit a chuckle. The jokes are predictably bad, but they often hit the mark in context.

Otherwise, I always have backspace, which is the lifeblood of anyone who writes for a living. It’s never a bad thing to take matters into your own hands. In generative AI workflows, this is more of a necessity than an optional caveat.

Another reason I love Gemini is that it saves time. Being part of a newsroom means things get a little too dynamic. To be honest, this is a pretty strange situation because AI is having a visibly negative impact on the media business.

In the rush to report on recent events, emails are often ignored, whether due to time constraints or simply because there is no energy left to keep up with inboxes by submitting articles in quick succession. did. That’s where Gemini comes to the rescue.

Now, I’m not a fan of email summaries, but they help me decide if I need to dig deeper into the actual content. When you open an email and choose a reply, Gboard’s dictation and Gemini’s smart compose features take control.

The entire pipeline is noticeably faster, and that’s what matters after all. This is especially important if you work across time zones, where 1 a.m. is when the news cycle and email onslaught are active.

Is that legitimate?

A healthy part of my work routine includes pitching to at least a dozen editors at a variety of reputable outlets each month. Unfortunately, I’m not alone in this fray. This means that your recipient’s inbox is always overflowing. And now comes the cruel part.

Few suggestions are accepted, but most emails end up in a black hole of no response. In the early days, I thought that any respected editor should have the decency to respond with, if not a polite rejection email, then at least a curt “no.” Realistically that’s not possible.

But I still feel the pain of the lack of response to the pitch I spent hours researching, writing, and fine-tuning. Interestingly, I’m in the same position as an editor, so it’s not a hard pill to swallow. At least a dozen times a day.

My inbox also includes an avalanche of stories, from a publicist asking an Armenian cryptocurrency evangelist to do a guest post to a company offering to review the latest safety smartwatches for kids. I don’t have any children. I don’t really understand virtual currency either. I laughed at the craziness of the situation and moved on to the next email.

Still, I can’t help but feel that the sender deserves a response from me. Approval or polite refusal. After all, it’s about basic courtesy, basic professional behavior, and respect for human dignity.

But if I were to carry that righteous burden and sit down to write emails, I would be wasting precious time and energy that could be spent on real work.

But if I used an AI agent to do a human job, would it hold any meaning? I decided to avoid that dilemma and get the job done. I’m currently drafting an answer using Gemini as my agent.

More precisely, I need to refine my coarse language and take on the burden of linguistic precision in my responses. But ethical questions remain, and depending on who you ask, the use of AI in communication can lead to vastly different views.

I’m in a similar dilemma and often re-read this passage from the Journal of Business Ethics to help me feel better. It’s titled “The Ethical Implications of Artificial Intelligence (AI) for Meaningful Work” and is well worth a read.

“When AI takes on simple or complex tasks that workers find boring or repetitive, it frees up time for workers to build autonomous capabilities by doing other more difficult or serious tasks. For example, AI could prioritize employee emails so that they only see the ones they need a response to, allowing them to prioritize others more valuable. You might be able to work on a task.”

big picture

The summary here is that if my use of AI speeds up the more mundane parts of my job, leaving enough room for me to engage in more meaningful work, then it is justified. That’s it. But this is a two-way street, and the person on the other side of the loop may not share those opinions.

Hussein Arnis Khan, an Alex Chernov Scholar at Melbourne Law School and an expert in legal system design and regulation, says that rather than long, flamboyant emails, emails that are short, typo-filled, or even brusque He said he prefers replies. human effort and ingenuity.

“What’s missing is the human touch. That’s what most of us want or desire,” Khan tells me. His attitude softens when I tell him that I’ve trained the Writing Assistant (also known as Gem, according to the Google AI Dictionary) on my writing samples so that every email reply sounds just like me. I did.

A paper published in the Journal of Autonomous Intelligence also highlights the importance of the proverbial human touch. It has been hypothesized that the widespread use of AI apps will come at the expense of real human interaction, leading to a loss of emotional connection.

What if he never guessed that AI plays a fundamental role in our communications? , Khan said after I showed him how to adjust the tone of his emails to make them eerily similar to the jovial tone I use to joke with friends.

But here’s the real question. In my self-proclaimed ethical quest to speed up (and ease) my workflow, am I unwittingly undermining the other human being? Messages shaped by AI Are humans entitled to receive this?

“Ignorance is bliss,” I tell myself. Maybe I’m choosing the less harmful option of leaving someone as “read” or responding to their request. It’s a matter of where your conscience swings. Personally, I chose to be more humane in my actions.

We’ll give you the answer (with the help of AI). It’s not exactly human, but it ultimately connects humans with each other in a solid two-way conversation. In the end, you could say it was a win.