🎉 The SWE fantasy is dead 🎉
and we killed it with thinking machines
Hi! Thanks to those of you that came to the 1st Engineering Engagement Showcase (EES)! If you missed our sermon, here it is in all of its glory.
The Future of Tech
5 years ago, it was a great bet to study hard at a top university cs program, secure the best internships, and graduate with a nice cushy offer at some FAANG company. Earning 200k or more straight out of undergrad, you could expect to work only a few hours a day and retire at 37 in the Bahamas.
Not only is this no longer the case, it never will be ever again. The world is changing, and so is the future of every single industry, including software engineering. Layoffs are only the beginning.
It is no secret that the tech market is in a free fall. Over the past year, over 200k tech employees were laid off, more than 50k in the past 3 weeks alone. Now, many may argue that this is simply COVID overhiring corrections or just a temporary setback. “The market will bounce back, it always does”. While the latter half is true, the jobs will not follow. The SWE fantasy is dead, and we killed it with thinking machines.
The fact of the matter is, the technology to replace the majority of software engineers is at most a year away, if not already here. While systems like ChatGPT are not perfect, they prove something. Programming is a solvable problem and we have already solved it. ChatGPT is a simple extension of GPT-3, built back in 2020. There are now internal models more than 3 times as large, not to mention GPT-4, which is coming out next month. Just like chess, go, and protein folding, AI will and has already started consuming the field of computer programming. The only difference is now, an entire industry is on the chopping block.
Bojan Tunguz, a research scientist at Nvidia and 4-time Kaggle grandmaster, has said, “AI is not the next big thing… [it] is the only thing.”
AI has progressed far beyond being a tool for software engineers to make web development smoother and is now transitioning to a source of automation. At this rate, it is quickly becoming a question of what can we do for machines, not what machines can do for us. These systems are continually outclassing us, and soon they will outclass us at everything.
The Future of Everything
My good friend Luke introduced me to waitbutwhy, a blog that breaks down complex topics into short, captivating articles. Here is a graphic from one of their pieces on the AI revolution.
We are right here, if not along the path already. This is all due to the borderline daily advancements in artificial intelligence. Advancements that only a small group of people know about and have the ability to iterate upon. Advancements that will not only transform the job market but the course of humanity forever, lead by a small group of researchers that are fully aware of the path their field is on. Strive to be a part of it.
“the other 40% [of hired contractors] are computer programmers who are creating data for OpenAI’s models to learn software engineering tasks.” - OpenAI has hired an army of contractors to make basic coding obsolete
6 years ago, the greatest achievement in AI was when Deepmind’s AlphaGo beat a grandmaster at his own game. Many thought that we did it, this is as far as we go. Now, anyone with a laptop and access to Youtube can build a system on the same level. Hell, we have machines that can pass the bar exam. Not to mention cicero, an AI system that beat some of the best players in the world at Diplomacy, an online public relations simulator, and arguably the most human game we have at our disposal.
Finale
It is important to understand that we are on the brink of a new era. As we rapidly approach the point of artificial general intelligence (AGI), where machines surpass human cognitive abilities, the future of AI is shifting from a focus on technological advancements to a struggle for power. The power to train systems to conquer any task. AI is quickly becoming not only the field of the future, but the field of everything.
With the ability for computers to become experts in any field and invent new technologies at a faster pace than humans, the advancement of humanity as a whole will accelerate at an unprecedented rate.
Very soon, we will solve intelligence, and solving intelligence solves everything. Nothing else matters.
Cool Down Lap
Reining it back in for a second, is the goal of our organization solving intelligence? Of course not, we will leave that to the researchers making millions a year at OpenAI and Meta, for now. We want to be a part of the conversation. We want to contribute to the most important field, and the time to start doing so is yesterday.
AI-Generated Image of the Day (“my mind is ranking the reality up and down”)
