I am a professional writer who reliably completes multiple articles in my day job, and in my off hours I am immersed in writing and can tackle any writing assignment And yet, I would rather submit a 1,000-word article on deadline than write a simple text message

I don't know myself what it is about text messages, but text messages are not my preferred means of communication I struggle to find the right tone I vacillate between a straightforward response and a lengthy reply that would be better suited for email I struggle with nuance and wonder if it is appropriate to respond in a joking manner or if I should respond in a straightforward manner

So perhaps I am a supposed reader of the chat assist feature in the new Galaxy S24 as part of Samsung's push for AI-powered features

If you are not familiar with this particular Galaxy AI tool, Chat Assist can assist with text messaging in several different ways With the tap of a button, Chat Assist can check the spelling and grammar of your texts and even translate them into different languages The most notable feature, however, is Chat Assist's ability to take a message you have written and reconstruct it into a different writing style - more formal or polite if you are texting a work colleague, more causal and emoji-filled if you are chatting with a friend

It's all very simple, as I show in my guide to using the Galaxy S24's chat-assist feature In other words, will the person I am texting suddenly suspect that my message is not from my mind, but directly from the neural processor of my smartphone?

To find out, I began sending messages to a collection of family, friends, and colleagues to see if they could distinguish between authentic Philip Michaels texts and texts massaged into text form by the Galaxy S24 Hopefully, we'll see how effectively Chat Assist's text can be distinguished from our own, and maybe we'll be relieved of the stress of chatting

"It's easy to tell which texts are yours," my wife told me when I approached her with this experiment 'Anything that isn't a monosyllabic message is obviously written by a robot' I must confess, she is keeping tabs on me

Nevertheless, we got to work on the first test message In it I attempted to make dinner plans and tried to trick my wife into thinking I had manners by using a polite rewrite of the chat assist My wife easily saw through the ruse and correctly identified the first text as coming from me

I decided to include an emoji to make her realize that the message was more affectionate (Primitive people use pictures to communicate We have letters) Alas, the chat assist added an emoji and the jig was over

By the way, you should always send a little email to someone special to let them know you are thinking of them, but don't leave it to a formal suggestion from Chat Assist 'Greetings, I have ensured your well being I hold you in the highest regard"

Oops, Chat Assist, don't be too thick

Our last message covered more mundane chores, and Chat Assist's polite option gave the game away by turning "I need you to pick something up" into "Do you need to buy something?" My wife once again saw through the processed text and identified my actual message perfectly 3-3 It was nice to know that someone had been paying attention to my messages for the past 20 years

I have known Brian for 10 years, so he should have some idea of what text messages I am sending That said, Chat Assist is probably better suited to assuming the casual tone that one might expect in a conversation between friends

Chat Assist successfully tricked Brian into thinking the AI-generated message was from me by rewriting in a casual tone the Cal football predictions that would surely come back to haunt me in the fall Words like "recon" and phrases like "what's your take on that" seem casual for a human to write

Like my wife, Brian knows that I am not the kind of person who uses a lot of emojis So choosing the "emoji-ize" option for the chat assist to try to re-edit my Magnum detective-themed text message was probably the wrong choice if I wanted the AI-generated text to appear as my own

I tried the polite option to invite Brian to lunch Do I really use high pejorative terms like "convene" or phrases like "Please specify your preference" when chatting with friends? Apparently so

I've been working with John Velasco of Tom's Guide for over a year now, and we are in constant contact since we are both in charge of smart phones And since the two of us are on opposite sides of the country, texting is our primary method of communication John would probably find it better to tell the difference between chat assist messages and my messages

John is writing an article comparing the Galaxy S24 cameras, so I thought I would send him a congratulatory email However, Chat Assist's attempt was a bit too enthusiastic and included some emoting for me to trust

Warning employees about coming to work without pants: this was just a thought exercise, and through Chat Assist it got a little too formal Again, John spotted the real text

A typo in my original message about going to get lunch for John confused him, and I made the confusion worse by intentionally including an emoji in the original text Chat Assist removed that emoji, but casually inserted a playful "yo" Still, the emoji and typo fooled John, who fell for Chat Assist's tampered message [I have worked with Anthony Spadafora a bit, but not enough to be in constant contact with him I figured it was enough to simulate an email exchange with someone I recently met at a party or work event In these relationships, it is particularly difficult to strike the right tone, and the chat assist is likely to be most useful

For this test, I also changed my approach Instead of sending Anthony the actual text and the version edited by Chat Assist, I sent him the four tests alone and asked him to identify which were real and which were AI-generated From top to bottom, the first and last texts were created by me, and the middle two were modified by Chat Assist

Anthony determined that there was no way the chat assistant could have included that emoji and correctly deduced that the first text was written by me (Oh, if only he knew) He was also correct when he guessed that the second text was composed by the chart assist

However, Anthony guessed wrong about the third tweet because he thought the AI did not use "fave" No offense

I think Chat Assist is most compelling and effective when you are texting someone you are not that close to and you are not sure what tone to use in your message It can also be useful when creating formal texts to colleagues, but depending on how closely you work together, it may be easier to find texts created by an AI

My general observation is that the more formal and discreet your texts are, the more work-oriented Chat Assist is If you try to tweet casually, you will either be easily spotted or it will not fit your usual tone

Chat Assist is not necessarily my favorite Galaxy AI feature The ability to use AI to reformat and summarize notes would be much more useful However, if you feel like you don't know what to say in a text message to someone who is not your lover or close friend, the chat assist feature could save you from sweating over speaking in the right tone