How to write emails using AI tools, This does not imply that someone got into their selected career because he or she was fiercely in love with the management of in-boxes. But there, at your job as a creative director, a software engineer, or even a mid-level manager, we spend a dreadful part of our lives.I do still recall one specific Tuesday last year. I opened my laptop and saw a bunch of 14 messages concerning the creeping project scope, three of which were labeled as urgent on matters related to clients that did not bother to factor in my office hours, and at the same time a confidential message concerning a vendor and their pricing.
Over the last 18 months, I have been using AI writing tools daily, and the flashing cursor is all that I used to see before I could have a second cup of coffee.Now, I do not see them as a trend anymore, but as a junior assistant that I have to have. I have found that you do not need AI to do everything. It is about conquering the resistance of the blank paper without letting your hands move from the steering wheel, using AI.An experimental working principle of composing emails through AI, but not like a soulless robot, is the following:
The “Sous-Chef” Philosophy
The biggest mistake I see that people committed is clicking the Send button after pasting an AI response in the Outlook or Gmail account. Please, do not do this in the name of your reputation. AI is not a ghostwriter it is your sous-chef. At the workplace the sous-chef is slicing the onions, boiling the stock and setting up the station. But it is the head chef (i.e. you) who tries the sauce, season and serving of the dish. When you work as a prep cook, your emails will become efficient because of AI tools. Your emails addressing them as head chef are generic, hallucinatory and simply too enthusiastic.
The Rough Draft: In this step, one needs to overcome the paralysis of the draft version (Step 2). The toughest part of a challenging email, the emotional burden, has to do with the first sentence, e.g. declining a job offer or writing about a missed deadline. This is what I use AI to alleviate this tension. I enjoy the chat interface which I use as a dictaphone. I ramble.
My Prompt:
To one of my clients, Sarah, I should write an email message. The API integration is erroneous and we are not going to meet a Friday beta launch date. I have to make us get to next Wednesday. Be professional but firm. There should be no excuses; one should simply provide facts and a solution. The AI spits out a draft. Usually, it’s a bit stiff. It can make use of sentences like I am sorry to tell you (too formal) or getting into the problems ( archetypal AI tell ). But the secret is being able to have a framework now. I do not see a white screen anymore. I have 200 clay smears to model. I will be able to svelter the fluff, smoother the mouth, and cut my voice. This process would turn a 15 minute panic attack into a 3 minute editing project.
Step 2: Context is King (Garbage In, Garbage Out)
Reason being you are not providing your AI tool with context when you receive generic emails. I have also learnt that there is a difference of enormous gap between the relationship dynamics in your prompt.
Bad Timing:
“Follow up email with John. Good Prompt: This is a short informal follow up message to John. Be cordial, we have a history of five years together. Enquire in relation to whether he had a chance to examine the contract which he was sent last Tuesday.
No fluffy intros.
” The relationship (5 years, friendly) and tone constraint (no fluff) in the request transforms the response of my request to display the text Hey John, Hope you are doing well, rather than the first message of my request, which was, Dear Mr Smith, I hope this email finds you well. Were you able to study that Tuesday contract?
Step 3: The “Tone Polish” of High-Stakes Comms.
On some occasions, I write the email, and my feelings are bleeding in the text. We have all experienced this experience: one of us has received a passive-aggressive letter, and you have typed a stinging reply. You are not supposed to send one, yet you have to make the point without provoking a war. This may be considered one of the most appropriate uses of AI.
To do this, I will paste my infuriated and disheveled writing to the tool with the instruction: “Write it again, so that it is diplomatic, professional, and collaborative. Keep the nature, eliminate the violence, which is that we simply cannot countenance this budgetary cut. The instrument is a kind of emotional filter. It takes away the sarcasm and proceeds to the business justification. It has saved me on a number of occasions where I am planning to burn bridges.
The fourth step is the Human Sandwich Technique.
So you can stay influential and high in terms of your EEAT (Experience, Expertise, Authoritativeness and Trust) in what you say, I recommend the Human Sandwich technique I use. Top Bun (Human) The shaking and the hook of the introduction must be yours. Re-creation of a conversation, a joke that everyone knows, or something that an AI is unaware of.
Thanks to meeting you in the expo at Austin. Meat (AI-Assisted): Most of the information, dates, logistics, bullet points and summaries can be texted or designed with the help of AI. It is so convenient in recapitulating long threads or in arranging haphazard ideas in categories. Bottom Bun (Human): The sign-off. Don’t just use “Sincerely.” Use a sign-off that is appropriate to you.
You have to insert the AI content in between the human touch, or the recipient will know that something is not right as they intuitively do. A sixth sense of synthetic text is growing. There is a loss of confidence in the event that email is synthetic.
The Danger Zones: Ethics and Privacy.
We must sit, and have a grave consultation on the hazards. This is my golden rule as a person that enters into contracts and proprietary methods: Do not feed a publicly available AI model with confidential information.
This is because should you write an email concerning a sensitive HR issue, a trade secret or not yet published data about financial information, you must first do a redact and put it in a prompt. Add in such place holds as: [CLIENT NAME] or [REVENUE FIGURE]. Further, the verisimilitude is within thyself. This was one occasion that the AI has written a response concerning the software refund policy.
It audaciously developed a policy, which was not present. By sending that I would have been bound by law to make a promise and break it. Always fact-check the output. Check it out in case the AI gives a date, price or policy. You are the one that works, as a professional; the generator is a simple computerized text generator.
Output Analysis: What to Scrap.
You need to be a ruthless editor to have your emails sound human with the help of AI. It is a list of what I delete about the first thing that I write in AI drafts: I wish you a good morning.
How are you doing? (Overused).
“In today’s digital landscape…” (Filler).
“Is ecstatic, is rapturous, is Breakthrough. (AI tends to be hyper-positive. Communication in businesses is in fact less partisan. Unprofessional sentence patterns. (The cadence of the three sentences is inclined towards the same line. Mix it up.
The Future of Your Inbox
It is not unlazy to automate the emails. It is something to do with cognitive load. By writing an email to leave the heavy structural lifting to the company, I will be able to focus my intellectual power on what, really, requires my experience: strategy, creative problem-solving as well as the ability to make real relationships. By this, it is not having your personality automated. It is to clear the mess so that your personality is able to shine through the noise.
Start small.
This is employed to summarise a long thread. At this point prepare a meeting agenda. In no time, you would become a master of your inbox and not a victim of your inbox.
FAQs
Q: The question is that will AI have the ability to compose the emails in my specific voice?
A: Yes, eventually. Training is a process that analyses your style using an example of one of your past emails and then you can tell it to write in that style.
Question: Is it rude to compose emails using AI?
A: Not if you edit it. Particularly when you write a full wall of text without editing it, you are disrespecting the recipient of the message. It is a business etiquette to have a clear and concise method of communication.
Q: What is the most appropriate AI instrument to use pertaining to email?
A: It depends on your workflow. Large chat models (ChatGPT or Claude) work well when it comes to overall writing. In order to fix grammar and tone immediately, it is more closely attached to browser extensions (like Grammarly).
I: What will help the AI to sound less formal?
A: Have you some adjectives on your prompt. The default level of formality will be broken with such words as conversational, brief, punchy, or text-message-like.
Issue: Does one have to disclose the fact that I used the AI to write an email?
A: Generally, no. You make no reference to the fact that you utilized a spell-checker or thesaurus. However, in the case of the email that consists of AI-generated information or summaries, it is also possible to act ethically and check the information and point out that the summary was produced with the help of AI to accept the fact of potential lack of nuances.