AI Unplugged: We're Gonna Need More English Majors
Michael Tresca
Director, Marketing & Communications for Global Talent Acquisition at GE Vernova
When I decided to become an English major, I had two career choices: teacher or journalist. After my first business internship, I quickly learned that education does not equate to eloquence. That's when I decided to go into business communications instead, a field that I imagine teachers and journalists probably aren't very fond of. But there's a new job for English majors on the horizon, and for lack of a better term, I'll call them "AI Whisperers."
I'm Sorry, What?
Yes, you heard me (even if I was whispering!). AI Whisperers, also known as prompt engineers, are defined as:
...a specialist who excels in comprehending, communicating with and guiding AI systems. Equipped with a rare combination of technical acumen, psychological understanding and communication prowess, AI whisperers serve as the bridge between the human and machine worlds. The primary responsibilities of AI whisperers include interpreting AI behavior, identifying the strengths and limitations of AI systems and directing their development to achieve specific objectives. AI whisperers collaborate with domain stakeholders to ensure that AI systems operate in harmony with human values and ethical guidelines.
Well that's one definition anyway. The field isn't that new, but now that AI has been launched en masse to the general public, AI Whisperers are becoming very important. AI Whisperers whisper in someone's ear -- and in this case, that of AI -- that they got something wrong, or need to get something right. They are the secret sauce that makes AIs more human-like, and we're going to need a lot more of them to train these digital creations. Unfortunately, as is often the case with tech fields, the value of communication is sometimes minimized.
When the Internet became ubiquitous, there was a brief, shining moment where I was convinced English majors would rule the world. Surely, those of us most proficient with the English language would thrive in a space where text and communication rule? But then came do-it-yourself blogs, social media, and memes -- slowly but surely degrading communication to simple, easy to understand concepts rife with spelling errors. It became clear that as much as English majors were hoping to be the leaders in the digital space, we're often considered something of a nuisance, correcting people's grammar and arguing about possessive commas (see what I did there?).
You'll note that AI whisperers require creativity, social and emotional intelligence, and communication. These are all things English and Communication majors have in spades.
Why This is Important
The Poetry Test by Sierra Elman illustrates why these skills are important for the future. Elman tested OpenAI's ChatGPT-4, Google's Bard (at the time before its transformation into Gemini), and Anthropic's Claude-2 by surveying a panel of 38 AI experts and 39 English experts to see if they could detect AI-generated poetry.
领英推荐
At the time the test took place (January 2024), the AI were largely unable to fool the majority of respondents (nearly 90% figured it out). But the bigger story is that the test was as much of AI as it was of us. English experts were able to correctly guess the author of poems an average of 3.13 out of 4 times, compared to AI experts of 2.61 out of 4 times. They were more likely to get perfect scores too -- while 18% of respondents correctly identified the author of all four poems, English experts were scoring perfectly 28% of the time vs. their AI counterparts (8%).
Job descriptions for AI Whisperers seem to vacillate between "technical experts who have language skills" and "language experts who have technical skills." The Poetry Test points to the latter, arguing that we'll need more, not less, English majors to address AI concerns:
In the short term, I think that having professional writers help work on AI teams to improve future versions of AI has real benefit to both writers and the AI community. Better versions of AI free up writers to develop richer, more creative ideas and projects. It will allow them to more efficiently edit their works. And in the longer term, I am cautiously optimistic that AI will open up new paths and opportunities for creative work. What exactly this will turn out to be will be constantly evolving (and deserving of its own separate study). But it’s important for creatives to be flexible and able to pivot, and isn’t that already part of the DNA of the creative mind anyway?
Don't Give Up Hope
My Master's thesis in communications was about the potentially corrosive effects of anonymity and the Internet. Most of my instructors didn't know what to do with me, and were confused as to why I would want to write a paper about Internet anonymity when that seemed to be a technical domain. My first job after earning my Master's in in the business world was as a programmer creating HTML web pages (if you can call HTML "programming"). I left that job as a contractor ... only to be hired a year later by the same company as a communications specialist, doing the exact same job. Or to put it another way, what started out as a purely technical discipline became engaging enough to be an artistic discipline instead.
AI is at that point now. Learning to talk to AI, understanding its speech patterns, testing its language limits, are all skills that a human who is highly proficient in English and communications can provide. According to ZipRecruiter, the salary ranges from $20 to $64 per hour, or from $65,000 to $132,000 a year.
English majors are in free-fall in the U.S., declining by 30%, with humanities enrollment overall declining overall by 17%. The reasons for this vary, but they are as much about the perception of how much an English major will make it "in the real world" as it is about whether English majors are really "English majors" at all, but rather communication and emotional intelligence specialists with the skills to navigate a complex world. Those just happen to be the same skills that make AI Whisperers so critical to the development of digital platforms.
There's opportunity, potentially a lot, for English and communication majors. As AI gets better at mimicking us, we'll need those proficient in language to help edit them, guide them, and yes, even detect them. Now we just need universities and high schools to get on board.
Please Note: The views and opinions expressed here are solely my own and do not necessarily represent those of my employer or any other organization.
Head of Sales Development at Hightouch - Ask me about composable CDPs
7 个月Sending this to my dad and telling him I didn't waste my college years.
Accounting Manager at Gasket Resources Inc.
7 个月Absolutely true. Being able to communicate effectively and understand the nuance of language never goes out of style.
Driving Value through People, Process, Platforms, and Performance
7 个月Couldn't agree more, Michael Tresca. There are a few more considerations as well, as we move from large-language modeling to something like SSSMs, and a certain kind of English major will likely start to become more highly sought after. Furthermore, when the context of navigation changes from being machine-centered (structured interfaces) to human-centered (conversational using AI agents), the need for improving the understanding of context from the questions asked becomes critical. (Sort of like watching the speaker in William Blake's The Tyger https://www.poetryfoundation.org/poems/43687/the-tyger come to realizations through questions asked). In my mind, it will be those who *not only* can (as you captured so well above) manage the translation of context with how humans encode emotion into language (eventually the work of computational linguistics will weaken this need) *but also* can navigate what is not written or not said (which has been a part of understanding the nuances of semiotics and post-structuralist approaches in how they help people recognize what is said/written and what is not said/written).
Digital Candidate Experience Strategy
7 个月I love your take on this Michael Tresca and you're absolutely right. Last year, I worked on a small exploratory project to test an AI tool. While most of the time, given the very specific path we were asking the AI to follow, it worked. Other times, it took some really odd turns. The work is continuing to improve the experience and the accuracy but it's requiring a number of talented individuals to bring this tool to it's true potential.