The Age of Transformation

We Are Living in The Greatest Time in Human History + Google's AI Controls Robots

Welcome to edition #22 of the “No Longer a Nincompoop with Nofil” newsletter.

Subscribe 🖌️ | Go Premium 🚀 | Work with me 🤝 | Discord 🤖

Here’s the tea ☕

  • GPT-4 changed, but didn’t get dumber 🤖

  • The future of superconductors could be here ⚡

✨ From our Sponsor ✨ 

Streamline workflows and 10x productivity!

Text Blaze is a tool that helps people eliminate repetitive typing with keyboard shortcuts. With Text Blaze, you can create smart templates that work wherever you work.

GPT-4’s evolution

Last week, I referenced a research paper that claimed to show how GPT-4's abilities had drastically decreased from March till now. Turns out that paper was complete and utter garbage. The methodology used in the paper to test GPT-4 was ridiculous. In one instance, a response of GPT-4's was incorrect, but that was because the code started with random python markdown. This is one of the reasons why I generally don't talk about extremely new papers, not until I know for a fact that they carry substance.

As I say that, what I’m about to cover is another new research paper. I know what you’re thinking; why on Earth would I do that given what I just said. Well, you’re about to see why I can’t help myself. If the results of this paper turn out to be true, let’s just say, the world will never be the same again.

What even is going on?

We are incredibly lucky to be living in an age where medicine, transport and the general quality of life is so high; in fact it’s the highest it’s ever been. It shouldn’t be a surprise then that it’s also the greatest technological era of human history. Only very recently, with the rise of LLMs like ChatGPT did we essentially enter the next technological revolution. AI is taking the world by storm, and for very good reason.

What would be baffling then, is if another world changing technology would be founded during this period of time, especially so close to, and during the AI revolution. Well, we just might have the next big thing.

Researchers from Korea recently published a paper in which they claim to have found the worlds first “ambient pressure, room temperature superconductor”. If you understand the implications of this then you’ve probably already heard of it. If, like most of the world, you’re not quite sure what I’m talking about, let me break it down for you. It’s a wild ride.

Unlocking the next frontier

A regular conductor is a material that allows electricity to flow through it, like aluminium or copper. The issue with these is that as electrons flow through the material, there is resistance which causes heat; this is why our computers have fans and heat sinks.

Superconductors are materials that allow for the same flow of electricity, however, they don’t have any resistance. The issue with these is that they only work in extreme conditions, like extreme cold or pressure, like at the bottom of the ocean.

LK-99, named after two of the authors, Sukbae Lee and Ji-Hoon Kim, supposedly works in ambient pressure and at room temperature. I know what you’re thinking; is this legit? Well, there’s a few things to consider. There are actually two papers released for LK-99. One with six authors, and one with just three. Why? Because the Nobel Prize can only be shared between three people. The authors of this paper genuinely believe in what they’ve discovered, and they’re not some nobodies either. They’ve been at it for over 20 years and even patented the tech in Korea. But I didn’t wait over a week to tell you this.

The reason I’ve waited is because people are starting to attempt to replicate it, and well, we’ve seen success. Sinéad Griffin, a scientist at Lawrence Berkeley National Laboratory (LBNL), seems to have found evidence to back the original claims also. I’m not going to even bother going into the details of her work and make myself look like a fool, I will link multiple things in the short digest below.

There is so much happening surrounding LK-99 and the only place I’m seeing it is on Twitter. There isn’t even any media coverage I’ve seen. It is astounding considering this is literally world changing stuff that is also seemingly being proven as every day goes past. You want to know what people think about it’s feasibility? All you have to do is check the betting odds.

In the beginning the odds were incredibly slim; now it’s practically 50/50. The irony is that this is probably how the entire NASDAQ might be moving soon. If this is what we think it is, we will have to go back and redo almost everything. Energy grids, electricity, phones, batteries, rail; the foundations of modern society. The enormity of these discoveries cannot be overstated. One of the craziest things about all of this? It very well might be the case that the best place to mass produce these things is in space. Insane things could happen in the near future.

Also, if you consider this entire situation from a geopolitical lens, it changes a lot. Taiwan and South Korea, leaders in chip manufacturing would be seriously affected by this. This could level the playing field entirely.

Even if LK-99 is not replicable, we are already learning so much about superconductors and the processes surrounding their creation. The fact that such a monumental discovery could coincide with the rise of AI and perhaps even the creation of AGI, is truly an incomprehensible scenario. It is hard to fathom how different the world will truly be by the end of this century. This decade is shaping up to be the most transformative in human history… so far.

This is, without a doubt, the most exciting time to be alive.

Robots of the future

I know. How many things that are happening can be “the future”. Well actually, the things I cover in these newsletters is like the tip of the iceberg compared to what’s actually going on. The premium ones have most of everything I read, which is just an absurd amount. Regardless, this is something I had to discuss.

Google recently showcased their new model RT-2, which uses AI to control robots. RT-2, as Google describes, is a “vision-language-action” model. What does this mean? Let me provide some context.

Previously, the main method of training robots to do any task would be to give them a set of instructions and have them repeat the task over and over again, and tweak the instructions until they were just right. Naturally, this was really slow and it meant you could never use a set of instructions for more than one use case. Not very scalable and efficient. Thats where AI comes in and sets everything on fire.

We’ve had to reconsider our entire research program as a result of this change… A lot of the things that we were working on before have been entirely invalidated

Vincent Vanhoucke, Head of Robotics at Google DeepMind

RT-2 allows a robot to interpret instructions, create a step-by-step plan to carry out some action and then actually execute the plan. This has never been possible before. The model has the ability to actually tell the robot how to move, or rather, similar to ChatGPT, it learns to guess how the robot should move by translating the robots movements into tokenised data.

Essentially, the AI is interpreting the world, and then speaking robot to the robot, and explaining what it should do. The applications of this in offices, homes, warehouses are vast, and with AI powering these machines, humanities adoption of robots will come sooner than expected. (I’m afraid to even think of the military applications)

Short Digest

Enjoy these points for further reading and exploration. If you want massive amount of information like this, consider going premium and supporting me write these newsletters.

  • Threads explaining the LBNL paper [Link] [Link]

  • Implications of LK99 [Link]

  • Sinéad breaking down her paper [Link]

  • Thread showing a list of people to follow and things going on [Link]

  • Accounts to follow if you want to stay up to date with info as well as all the drama. There’s a reason there are two papers.. there are serious disagreements between the authors [Link] [Link]

  • Thread explaining four papers discussing LK99’s potential superconducting abilities [Link]

  • This person is baking LK99 at home in their kitchen, albeit with different methods [Link]

There’s a lot more to add, be sure to check out the longer posts in the premium newsletters.

As always, Thanks for reading ❤️

Written by a human named Nofil


or to participate.