Did You Write this Email or ChatGPT?
Recently, I found myself in a perplexing situation after composing a crucial email offering feedback on a technological solution to a colleague. I approached the task with meticulous care as the topic mattered deeply to me and I dedicated considerable time and effort to craft my email, aiming to convey both precision and diplomacy.
Despite my firm will to preserving the authenticity of my voice, I sought assistance from chatgpt, in refining certain nuances within the email. Upon receiving my email, my colleague was pleasantly surprised by the depth and eloquence of the content. Given he knows I am in favor of integrating chatgpt in everyday life at the office for increased productivity and creativity, my colleague couldn't help but pose a question that encapsulates the essence of our evolving relationship with generative AI. He asked me directly, "Was this email yours or of chatgpt?"
This question has given me much to think about. It delves into the intricacies of authorship and the impact of AI assistance on our written communication. The boundaries between our personal contributions and the augmentations by AI have become increasingly nebulous.
I believe that the partnership of human intellect with AI doesn't necessarily diminish the authenticity of our expressions; rather, it amplifies the expressivity and efficiency of our communication, allowing us to convey our thoughts with clarity and impact. All this, provided that we take full ownership and accountability of every single word and message of the text. Ideally, AI helps us structure and articulate our own thoughts, not adopting and reproducing thoughts foreign to us. Those who do so risk a dissonance between their emails and their personality established through time at the office.
Reflecting on the not-so-distant past, the process of crafting an important written analysis often involved a visit to the local library to gather information and facts. However, the advent of Google revolutionized this process, enabling us to access a wealth of knowledge from the comfort of our homes. Subsequently, Autocorrection tools emerged as our companion, ensuring the quality of our orthography, syntax and grammar. Before the aid of these tools, these three were not just formalities; they revealed parts of our identities, our background and education.
领英推荐
With time, we assimilated autocorrection into our writing routines. Nobody asks us anymore whether our orthography, syntax and grammar are ours or of autocorrection tools. If we mention a fact, no one asks us whether we knew it from our personal education or we googled it. Thus, I believe it's inevitable that we will adapt to the assistance of generative AI in crafting emails at the workplace as well. We are in the early phases of a process that will eventually reach an equilibrium where its healthy and smart usage will be taken for granted.
The topic has many nuances, various arguments for and against the usage of generative AI for crafting emails at the workplace, along with potential gains, drawbacks, and risks. I always seek to learn from diverse voices. My opinions also evolve, especially on such a new topic, and my stance on it may change with time and experiences. What are your thoughts?
PS: Yes, you guessed it right! This post was aided by chatgpt. The final version is a synthesis of many passages of the chat, plus heavy personal editing and new content on top. Chatgpt is among others my favorite English language tutor.
PS2: Again you guessed it right! The image is from DALL E. Not good, but I have no time to optimize it ;)
Computational Linguist and NLP Specialist Principal Key Expert in Artificial Intelligence
12 个月Great article Dr. Ioannis Petrakis, I also believe that these models are of a great value for spelling and grammar correction. The important part is of course not to lose your guard and be tricked by fluency when it is used to generate new texts. While the output might be very fluent and articulate as soon as one starts actually questioning the value of what was generated one sees that behind the fluency hides triviality. Those models are great cliches generators, true creativity and novelty still come from humans.
Tech Hub at Siemens
12 个月Hi. So far I've only used ChatGPT (actually the internal one) in the aid of writing texts to "round" some of my replies. I tend to be refered as somewhat blunt (direct to some, blunt to others) and I've used it to write a polite version of my draft. If I use the AI version, anyone who knows me well would see that the text wasn't mine, so I always end up using a merge of my version and the AI. That way I won't "sound" so blunt
Crafting Audits, Process and Automations that Generate ?+??| Work remotely Only | Founder & Tech Creative | 30+ Companies Guided
12 个月Articulating our thoughts with the help of AI can indeed enhance communication when done with authenticity and accountability. Well said! ??
SIAM Strategy Manager at Siemens
12 个月The good thing is that currently it is still quite easy to distinguish between a physical person and ChatGPT text, especially if you know a person. ChatGPT uses Cambridge or similar level of English while using lots of words non-native speakers never heard of...