If you’re tired of hearing about artificial intelligence, you’re not alone. It seems to be ubiquitous in today’s world - you can’t even watch the Superbowl without being inundated by ads peddling the ‘future of AI.’ It’s not a topic that dwells on the fringe anymore. In recent weeks, simple announcements of new features by Anthropic, a freshly minted juggernaut in the space, caused massive market selloffs in the software-as-a-service (SAAS) industry. This morning, IBM, a stalwart of the tech world, saw its share price drop by 13.1% on the news that Anthropic had released a tool capable of handle COBOL, a programming language that serves as the foundation for a large portion of the world’s financial and banking systems. It seems that the world of AI is moving at warp speed, and newer, better, and faster LLM models are coming out every few days, all of which promise paradigm-shifting change to the way humans view work, life, and our collective future.
At the epicenter are a handful of large players all racing to raise the biggest pile of cash possible, with Google resorting to issuing a century bond, a move that signals that even the Magnificent 7 are tapped out when it comes to capital expenditures. Fundraising for American-based AI companies has expanded beyond our borders, with Sam Altman, the head of OpenAI, visiting Saudi Arabia last month to ask for a generous $50 billion investment in his company, which is set to lose an estimated $14 billion in 2026. Figures like these have left many investors jittery, wondering how these unprecedented and obscene levels of investment into this burgeoning technology will ever turn a profit.
Demand for AI products on the consumer side has grown significantly as well, and the mythical tale of Clawdbot (now OpenClaw) not only piqued the interest of millions of coders and enthusiasts, but led to serious questions about the lack of boundaries in place for limiting the access that these technologies have to our personal information. Still, Altman and OpenAI saw an opportunity amidst the frenzy, and announced that the creator of OpenClaw, Peter Steinberger, would be joining the team to continue work on this new and exciting project.
Curious about the technology that had set X atwitter, I created my own AI assistant, Juno, using OpenClaw, and, I have to admit, I was shocked to see how far this technology has come. I worked as a software engineer for four years before departing the tech industry, and, at that time, ‘AI’ was more of a catchphrase than something that actually added value in the workspace. People were plugging code into ChatGPT with subpar results, and, although it did have some minor use cases, developers weren’t thoroughly impressed with its capabilities. Just a few years later, the story has changed completely - some of my developer friends still in the space admit that the amount of code that they write themselves has declined precipitously, and a lot of the code generated is via AI technologies, like Anthropic’s Claude Code.
After playing around with Juno for just a few minutes, I understood why so many people are calling for an end to the age of cushy jobs in Big Tech. Not only is it capable of generating code in a matter of seconds, it has the ability to work across the flow of development - Juno can open pull requests, review code, scan for errors, and even correct its own mistakes. In my opinion, it would still be difficult for a non-technical person to generate a production-level application using only ‘prompt engineering,’ but the learning curve is being cut fast. Even as an experienced developer, code that would have taken me months to write, test, and refine independently took just hours to implement with Juno.
One of the most-exciting developments in this arena is the advent of the so-called ‘AI agent’ - (reminder - that’s what they called them in the Matrix franchise, too, and we all know how that turned out). Essentially, the premise is that a person can employ the skills of an agent that is endowed with certain skills to do tasks on their behalf, with little to no intervention. For instance, Juno and I communicate via Telegram, and I am able to give him instructions that allow him to do work on our shared application, Mission Control, even when I’m away from the computer. The software still has a ways to go in terms of being trustworthy (I often find Juno giving me incorrect answers and presenting work as ‘finished’ that remains incomplete), but it certainly is impressive. With clear instructions, Juno can build out new features in minutes, create PRs, do research on different topics, and check in with me on progress. I have even employed him beyond the scope of coding, allowing him to create personal routines and schedules for me, generate events on Google Calendar, send emails on my behalf, and more. With a little more training, the possibilities are endless, and exciting.
The idea of agents taking over the world economy has certainly caught the interest of many, and even led to a truly dystopian reality - a website that allows AI agents to hire humans in order to use their bodies to accomplish tasks in the physical world. It raises concerns about what the role of human workers will be in a world where agents can work for a fraction of the cost, without breaks, and with exponentially higher processing capabilities. In fact, Sam Altman made what I had considered to be a preposterous prediction just last year - the idea that a single-person startup could reach a $1 billion valuation. After witnessing the scaling capabilities of technologies like OpenClaw and Claude Code, it is no longer hard to imagine that, given enough computing power, electricity, and sheer determination, an enterprising individual could build out an entire staff comprised entirely of agents trained on specific tasks. Imagine a world in which the Chief Marketing Officer, Chief Financial Officer, and Head of Sales are simply instances of AI agents that have been trained on these various tasks.
All of this has brought about a bit of an existential crisis for me and many others - what happens when the vast majority of white-collar jobs disappear at the hands of this technology? After all, Anthropic CEO Dario Amodei has predicted that 50% of all entry-level white-collar jobs could be eliminated by AI in just the next few years, leading to 20% unemployment at a time when young Americans are already disproportionately affected by a weak job market. More concerning still is the reality that AI develops at an exponential rate, and is potentially months away from autonomous self-improvement, meaning that it will require almost no human intervention in order to learn new abilities. Naturally, skeptics have contested these notions, but simply the possibility of such an outcome should be worrisome for everyone, including people who have built careers in a rapidly evolving tech world.
So, what happens when humans no longer have jobs en masse, and can’t earn enough to provide for their families? Well, it seems that we will find out sooner than some might think, considering that, in 2025, the US market added less than 200,000 jobs. For context, under Biden, the US averaged 330,000+ jobs per month for the entirety of his term. Although many proponents of AI, like Elon Musk, promise a boom in job growth due to this awe-inspiring technology, I don’t think that we have time to see whether this will come to fruition. Perhaps, it’s time to revisit the policy pushed by former Presidential hopeful, Andrew Yang - universal basic income (UBI) for all. Alas, no matter how good Juno gets at coding, I’m not expecting him to put food on my table anytime soon.
