troymc
troymc on Hacker News
I think that quote was coined by AI researcher and robotics engineer Yacine Mahdid writing under the social handle @yacineMTB
Andrej Karpathy circulated it online and placed it at the center of his address at Sequoia Capital's AI Ascent conference in April.
The idea is very old though. There's a story of how King Ptolemy I Soter wanted to learn geometry from Euclid, but being a King, he asked for a shortcut to understanding it. Euclid replied that "there is no royal road to geometry". (A "royal road" was basically a fast road reserved for royals, such as the Persian Royal Road.)
Is that the same thing as "The Singularity"? If so, why the name change? The name "singularity" is useful, at least, because of the analogy to a break in the model after which prediction becomes impossible.
In my opinion, the coolest thing in NotebookLM is the podcast-episode-generator. Each one sounds like two people having a conversation. It's fun to listen to a podcast episode about some niche topic (e.g. nuclear isomers, or the Weyl curvature tensor) while I'm cooking or driving.
Fuchsia ended up in some Google products, such as Nest Hub, Nest Hub Max, and Google's smart speakers, thermostats and displays.
But Fuchsia won't be in the Googlebook because there's no Chrome browser for Fuchsia. (In early 2024, Google officially stopped trying to port the full, desktop version of Chrome to Fuchsia.)
On second thought, I think it's not for K12 after all.
Yes, you are probably right.
Now that I look closer at the Googlebook, I think you are right.
I think it's a successor to the Chromebook. In the vast majority of modern K-12 public schools, the school district owns the hardware, not the students.
I guess it will be running Google's new operating system (a "modern OS designed for Intelligence") that combines elements of Android and ChromeOS.
Edit: Probably Android at the core, and then a desktop-grade Chrome browser on top.
Aw, c'mon man, they're just doin their job, ya know? Chill out.
I can relate because my dad is 84 and he really struggles with simple things like entering a password to sign in to Gmail. He forgets what he did last time and so I'm back to explaining how moving his mouse causes the pointy-arrow thing move around on the screen, to get it pointed at the wide rectangle near the middle of the screen, etc. No UI library is going to solve his struggles.
I solved most of the sign-in problem for my dad by picking a simpler browser than Google Chrome, and by tweaking his browser settings to be just-so. That's not going to be much help for you, the website creator...
Maybe allow passkeys for login? These days, passkeys usually get stored/supplied by the underlying OS. (By usually, I mean that's the statistically most common source of the passkey today. They can also come from a browser plugin or a hardware key.)
Maybe bring a (printed) book, brochure, flyer, or treatise on the nocturnal behaviours of silkworms?
"OpenCode Go" (a subscription) lets you use lots of hosted open-weights frontier AI models, such as GLM-5 (currently right up there in the frontier model leaderboards) for $10 per month.
And no doubt on next week's episode of everyone's favorite podcast-within-a-podcast...
I've used it to create little one-off tools that I needed for some specific tasks, without any care whether it's "of substance" or "investable." In the past, I might have Googled around to see if there was an existing app or open-source package to do the job. Quite often, the AI agent will use some existing software packages, but I didn't have to find them and figure out how to use them.
A real example from today: I got tired of the Looney Tunes way-too-colourful screensaver options on my Mac mini, so I asked Claude how to get a screensaver that is a nearly-black uniform monochrome grey. Surprisingly, that's not actually an option in macOS System Settings. So Claude just wrote a little Python script to generate the images I need for my two displays, saving them to the right place on my Mac mini. It used the Pillow package, but I didn't have to spin up a whole Python project and install Pillow; it just used PEP 723 inline script metadata to tell uv to use Python >=3.12 and to install Pillow. Then Claude gave me the uv one-liner to run the script (uv run ~/make_screensaver.py) and instructions for how to tell my Mac mini to use the generated files. The whole process took about 15 minutes from when I started writing the first prompt to the time I had my new screensaver working.
The plaid trim on the official uniform definitely gives it a Scottish aesthetic.
For me in Canada today, Kagi is showing nanoclaw.wrongtld as the third text link, after two different GitHub repos (why two? I didn't have time to sort that out). I clicked the thing to block the link to the site with the wrong TLD; hopefully other Kagi subscribers will do the same.
Wow, very cool video... and from 2015!
You wouldn't use an LLM to solve a big Linear Programming problem, because it would cost way more than using the Simplex Method, and you'd be worried that it might be wrong.
Is it named Coolwulf? The home page doesn't say its name anywhere.
The number of snacks in the basket is a random variable with a Poisson distribution.
I think it's now part of their training though, thanks to Simon constantly testing every new model against it, and sharing his results publicly.
There's a specific term for this in education and applied linguistics: the washback effect.
One reason is that you can get sued for libel or defamation. Another reason is privacy laws (weak to non-existent in the USA, but strong in Europe, Canada, and elsewhere).
Another possibility is that more people start using open-weights AI models (which have gotten very competitive recently [1]) running on cloud computing providers, and the main competition becomes just another aspect of the cloud computing wars: AWS vs. Azure vs. Google vs. Everyone Else.
[1] As I write this, GLM-5 is ranked fifth place on the LLM Leaderboard at https://artificialanalysis.ai/leaderboards/models
For most cars driving through air, at sea level, on planet Earth, at normal speed, the drag force F is proportional to the square of the speed (v^2).
That's not exponential because the speed (v) is not in the exponent. In fact, it's quadratic.
Corollaries: The power required to push the car at speed v will be proportional to Fv ~ v^3. The gas spent over time t ~ energy spent ~ power time ~ v^3 * time.
Tip: If you're in dark mode, flip to light mode so that you can see the graphics. There's a toggle in top right corner of the site.
Human teenagers somehow manage to learn to drive with less.
One of my favourite reads from this past year was Infinite Powers: How Calculus Reveals the Secrets of the Universe by Steven Strogatz. It's a wonderful review of the history of calculus, including intuitive explanations of the basics.