Major disclaimer: this essay does not represent my views of my company. I genuinely f***ing love working for them and feel blessed by what I do.
The year is 2035.
It’s been over a decade since artificial intelligence systems like GPT-4, Claude 2, Bard 2, etc, were introduced to humanity by companies like OpenAI, Nvidia, and Microsoft.
Humans no longer work in the traditional sense.
Artificial intelligence has completely taken over; our sole purpose is to write like monkeys on a typewriter to feed these learning systems.
It is frightening, the insatiable hunger of large language models (LLMs), the underlying technologies powering chatbots like ChatGPT. Humanity has long exhausted its supply of text, data, images, videos, books, tweets, encyclopedia entries, Reddit posts, and more. Original data is now as valuable as the graphics computing chips that serve as the brain of these systems. So that’s what our purpose is now — our ability to be original is now simply food that feeds the appetite of these systems.
Gone are the sounds of coworkers collaborating. Instead, open offices and cubicles have been replaced by Nvidia graphics cards crunching away at vector embeddings with a low consistent hum.
This is humanity’s future.
Or is it?
My name is Jacky, and I am an AI engineer at a healthcare tech startup. I am one of the “first crop” of folks applying artificial intelligence systems like LLMs to solving problems at the workplace.
Many many many many visions and analysis’ paint artificially intelligent systems like ChatGPT as the incoming new overlords of humanity, especially in areas where we once thought AI could never touch such as writing and art. But let’s take a brief walk down memory lane and examine previous world-changing technologies that were once also heralded as harbingers of human obsolescence and replacement.
A look into the past
In the 15th century, with the advent of the Gutenberg press, thinkers believed that this invention would cause information overload, eroding people’s memories, and capacity to think. Now we know that the printing press unlocked a level of access to knowledge previously unseen, leading to an explosion of ideas, literacy, and knowledge.
The printing press amplified human expression
In the 20th century, television was predicted to kill the movie industry, hurt radio, conversations, reading, and the patterns of family living.
Spoiler alert - it did not. The television transformed the way we consume news, entertainment, and content. It did not replace the movie industry, entertainment, nor the radio. Instead, the brightest adapted to this new format of expression, and a completely new way to deliver media was born.
The television amplified human expression
And of course, how can we talk about world changing technologies without mentioning the internet. The internet may be the previous generation’s single most transformative piece of technology. Again, these were the headlines - encyclopedias were done for, libraries would become obsolete, people would never talk face-to-face again, and more. The Telegraph wrote “Twitter and Facebook could harm moral values”, that this generation “cannot form relationships”, and how social media uses even “causes cancer”.
While the physical encyclopedia, Encyclopedia Britannica, did discontinue its print edition in 2010, the concept of the encyclopedia simply moved to a new medium. Wikipedia is in the top 10 most popular websites in the world, which broadened access to knowledge, and continues to be a highly curated and reliable source of information. Libraries continue to be critical community centers, and adapted to the digital age by offering the ability to “borrow” e-books, which is a very popular way to read today. And of course, human interaction - more than ever, people crave human interaction, especially while working and collaborating, and years of COVID lockdowns. The internet has not replaced human interactions, far from it, it simply added dimension (quite literally, thanks to Google) in the way we connect with others.
The fundamental human need for real interaction has not changed, the internet simply added another way for people to interact, especially those that live far away from their friends and families.
The internet has greatly amplified human expression
It turns out, these technologies did not become a wholesale replacement of the essence of what makes us human, only amplified it, and made it more accessible to all. All of these technological revolutions brought change, not replacement.
Change, not replacement
The concept of “change, not replacement” can be extended towards the impending AI revolution.
Rather than AI replacing humanity, we will instead be amplified and augmented by AI, and we will continue to adapt like we have in all technological revolutions.
Therefore, our fears should not be at “being replaced by AI”. Instead, it should be centered on not adapting and consequently being left behind.
Therefore, our fears should not be at “being replaced by AI”. Instead, it should be centered on not adapting and consequently being left behind.
As an AI engineer, I've witnessed first-hand the transformation these technologies are bringing about even in the workplace. AI is not eliminating my brilliant coworkers; instead, AI supercharges them. They are able to be more productive, more creative, have fewer blockers, and are more efficient. Let me put it this way - AI is a tool, an incredibly powerful tool, but it does not replace my coworkers, it only makes them even better.
AI does not replace my coworkers, it makes them even better.
AI at the workplace
From personal experience of seeing the impact of AI at the workplace, one’s proficiency in harnessing the powers of AI is directly proportional to your knowledge, experience, and professional expertise. The more adept you are in your field and the more knowledgeable you are about a topic, the more effectively you can guide AI to achieve your desired outcomes. Your excellence further fuels excellence from AI.
Your excellence further fuels excellence from AI.
The marketing team will be amplified, not replaced
In marketing, we are using GPT-4, Bard 2, and Claude 2 to generate cold outreach copy using historically tested examples, building personas of potential customers and audiences to model against, enrich internal databases using GPT-4 and the internet, create outlines to get past writer’s block, maximize SEO strategies for our blog, and more. AI has boosted our marketing team, and made our small team deliver results of one many times its size.
The sales team will be amplified, not replaced
In sales, we will be experimenting with intelligent lead scoring, predict which leads may convert and thus help prioritize sales efforts, transcribe and summarize sales calls so every person in the company has insight to the real problems our customers have, provide insightful recommendations for cross-selling and upselling, rehearse against objection scenarios, and more.
The support team will be amplified, not replaced
In support, Customer Service Managers (CSMs) can train custom models on all previous support calls and provide customer service through automated responses, have all our support docs available as a chatbot, analyze customer feedback and sentiment to help support customer’s pain points more effectively, identify the most popular feature requests, and basically free up our best people from common questions to give them more time to handle critical and complex issues.
The product team will be amplified, not replaced
In product, AI will enhance product development by automatically analyzing customer feedback, perform market research, test go-to market strategy, and conduct competitive analysis to provide insights into what features or products might be successful.
AI can further model individual human behavior and have thousands of “product testers” in all verticals + targeted personas to provide feedback to your feature and product - and yes, this is a real use case of AI that was documented in a recent research paper called “Generative Agents: Interactive Simulacra of Human Behavior (2023)”.
The engineering team will be amplified, not replaced
Finally, AI already augments engineering teams in numerous ways, such as Github Copilot quantitatively improving developer satisfaction and helps them conserve mental energy. One senior software engineer reports “(with Copilot) I have to think less, and when I have to think it’s the fun stuff. It sets off a little spark that makes coding more fun and more efficient.” The research conducted also shows that AI tools for development helps developers achieve faster completion times, helps them focus on more satisfying work, and ultimately find more fun in the coding they do.
The fact is, AI makes those that are already good at their jobs, even better.
So, how does this story end?
You’re probably asking me right now — “Jacky, about that dystopian image you painted earlier, there’s no way it ends like that right? Just now, you painted such a rosey picture of AI. We are SOO going to be living in a tech utopia thanks to AI!!!”
And you are right — I don’t think the AI revolution ends in the dystopian way I painted in the beginning. But I really hate to say this, because it may still end in a dystopian way, just not in the way you expect.
The dystopian ending may be this — as the efficiency of the individual improves thanks to AI augmentation, management will want to squeeze more out of you, and to do more with fewer people.
As the efficiency of the individual improves, management will want to squeeze more out of you, and to do more with fewer people.
It only makes sense. More than ever, we are seeing that corporations want to do more with fewer people. We are seeing it in every single layoff that has happened in 2023. The corporate world is brutal, and AI will only mean management will want more out of us.
This, however, is not an AI problem. This is a corporate complex problem.
It’s dystopian, but honestly, we sort of already live in this corporate dystopia. One thing’s for sure - the need of modern management and leadership to squeeze just a few percentage points out of you may get amplified with the rise of AI augmentation.
I think it’s fairly clear that if history is to repeat itself - it usually does, we will be amplified by artificial intelligence, instead of being replaced by it. That is not to say things will be exceptionally rosy, because humanity’s intentions stay the same, and so does the corporate world. I think employees need to adapt and change, because more may be expected out of us, especially with aggressive cost cutting in corporate America specifically. With AI augmentation, the baseline of what’s expected of us will be raised.
I encourage everyone to learn more about AI and the LLMs available today like ChatGPT, Claude, Bard, and more. They are very powerful, and genuinely save my team and I so much time. AI has made my already brilliant coworkers even more exceptional.
Tools like ChatGPT has also helped me a lot in my personal life such as with brainstorming for this blog post, being a Therapist Lite (this is not professional advice!), expanding on ideas, unblocking my writing, become a better person, and more.
I hope you learned something, and if you are interested in more essays on AI, sociology, work, philosophy, and applying AI, consider subscribing to the Topography of Applied AI. It really gives me encouragement to write down what I think about when I am daydreaming. And I hate writing, so any encouragement would be much appreciated.