Links

Andrej Karpathy has a very well-done Intro to Large Language Models video on YouTube. As a founding member and research scientist at OpenAI and with a two-year hiatus working on Tesla Autopilot, Karpathy is an authority in the field. He is also good at explaining hard things. As a Kahneman reader, I appreciated the Thinking Fast and Slow analogy proposed at about half-length in the video: “System 1” (fast automatic thinking, rapid decisions) is where we’re now; “System 2” (rational, slow thinking, complex decisions) is LLMs next goal. Also, I suspect Karpathy’s intriguing idea of LLMs as the center of a new “operating system style” is not too far off from what will emerge soon. The final segment on AI security and known attack vectors (jailbreaking, prompt injection, data poisoning) is also super interesting. ...

Quoting Jan van den Berg: This weekend we learned that Bram Moolenaar had passed away at the age of 62. And this news affected me more than I expected. Like so many: I did not know Bram personally. But I’ve been using a tool made by Bram for more than half my life — at least weekly, sometimes daily. That tool is a text editor. The best one there is: Vim. ...

Imagine, for a moment, that you are walking along a dirt road in the seventh century Middle East. The sun is hot, the air is dry, your feet are tired. It’s been a long journey, by boat and foot, from your home in Constantinople to where you find yourself now: outside of the walls of the mountainous river city of Antioch. In the bright sunlight, you strain your eyes to catch a distant glimpse of the sight you’ve come all this way to see – and then suddenly, you do. A bright stone pillar, stretching as tall as a church dome with an unsteady-looking wooden platform; and atop it a tiny, bedraggled, flinty old man. ...

Donald Knuth challenged ChatGPT-4 with 20 questions and then submitted the results and his comments to Stephen Wolfram. The whole thing is fascinating in many ways and worth reading. Some remarkable quotes: Of course these are extremely impressive responses, sometimes astonishingly so; thus I totally understand why you and others have been paying attention to it. The most immediate impression is the quality of the wordsmithing. It’s way better than 99% of copy that people actually write. It’s definitely not like a Markov model that uses the most predictable way to continue what’s already been said. On the other hand there are surprising lapses there too, as typical of any large system. ...

Adorable, functional, often internet-connected desktop bots like those below are hand-crafted by Mohit Bhoite in San Francisco, California. It pleasantly surprised me that they’re built as a purely artistic expression. All sculptures (as Mohit rightfully refers to them) are unique and not for sale. Check them all out on his website (via).

Today I learned about a precious little macOS command line tool, networkQuality. The networkQuality tool is a built-in tool released in macOS Monterey that can help diagnose network issues and measure network performance. Usage: networkQuality -v Example output: ==== SUMMARY ==== Uplink capacity: 44.448 Mbps (Accuracy: High) Downlink capacity: 162.135 Mbps (Accuracy: High) Responsiveness: Low (73 RPM) (Accuracy: High) Idle Latency: 50.125 milliseconds (Accuracy: High) Interface: en0 Uplink bytes transferred: 69.921 MB Downlink bytes transferred: 278.340 MB Uplink Flow count: 16 Downlink Flow count: 12 Start: 13/05/2023, 15:04:13 End: 13/05/2023, 15:04:27 OS Version: Version 13.3.1 (a) (Build 22E772610a) It supports Apple’s Private Relay, offers some configuration options and allows setting up your own server. More info here. ...

I read a well-researched story about Redis and its creator Salvatore Sanfilippo, also known as antirez. I was already familiar with many details as I have been following him since OKNotizie and Segnalo, of which I was a user. At the time, as a user, I exchanged a few emails with Salvatore, whom years later I had the pleasure of meeting in person, as we were both speakers at several conferences. ...

Minimalist News is the first LLM project that excites me but in a nervous way. Quoting the About page: We only publish significant news. To find them we use AI (ChatGPT-4) to read and analyze 1000 top news every day. For each article it estimates magnitude, scale, potential and credibility. Then we combine these estimates to get the final Significance score from 0 to 10. And now the best part: We’ll only send you the news scored 6.5 or higher. Sometimes it’s 5 articles, sometimes 2, sometimes 8. And sometimes — none at all. But one thing is constant — you can be sure that you haven’t missed anything important. ...

A fascinating article surfaced on Nautilus last week. Mary-Jane Rubenstein, a professor of religion and science in society at Wesleyan University, shares her concerns about the technical strides and aspirations of Elon Musk’s SpaceX, the company’s mission to enable thousands of people to live on Mars, and the ethics of terraforming the planet to be more like Earth. What’s intriguing, though, is Rubenstein’s thoughts about the religious underpinnings of the United States space program and how even modern science is still hostage to imperialistic Christian ideas. ...

Pitchfork has a great piece on Sun Ra and his legacy. It’s worth reading if you’re a fan, even more so if you know nothing about him. But what Sun Ra had done, and done best, was reminding earthlings everywhere that he wasn’t mortal. He was a signifier of a life beyond the reality of this one. He was a visual reassurance of the presence of another world. He brought the cosmos to the streets, and, most importantly, he was a reminder that one does not have to subscribe to the status quo—musically, stylistically, politically, ideologically. ...