To answer the question posed in the header of this report, it’s apparent that using open models is indeed the approach offering the biggest savings, followed by smart model routing. Spending controls and context optimization also bear down on costs, but they don’t come close to the first two techniques in results.
AI is grown more than designed - it is, to first degree, the product of repeating a straightforward optimization step many times on a hard-to-imagine amount of compute.
At the limit, and also well before that limit is reached, if all you do is fix the bugs, the AI will learn perfect optimization of reward, will realize not to reward hack in the perfect test environments, then turn around and reward hack in the imperfect real world environments.
AIs don’t just repeat the same sentence patterns but also the same themes (memory is a favorite), names (Elara Voss, Marcus Chen), and underlying ideas.
They've got blog posts of we had to rewrite this whole thing because the performance was bad. If it was always hotspots that made your performance bad, you'd never have to rewrite the whole thing. So, we know that that doesn't work anymore.
What is the underlying hardware capable of doing at its theoretical peak? And then you measure the delta between that theoretical maximum and what you have achieved.
One of the reasons that you don't see hotspot optimization as a thing that really matters that much anymore and one of the reasons I advise that architecture and and not making bad decisions is much more important is because a lot of libraries already have been optimized for you that you might use.
If every uh software engineer knew to watch out for false serial dependency chains, things where they were creating series of dependent operations that could not be optimized away or other sorts of architectural problems like that that cannot be easily fixed, then the world would look more like just wait and optimize the hotspot, right?
Now you can have a quake style console in your operating system. Except with an infinitely patient agent to help you customize it instead of esoteric variable names and slash commands like in the 90s
Smalltalk Best Practice Patterns. True, Kent Beck is better known for his later work, which is excellent too. But Smalltalk Best Practice Patterns is particularly strong on coding style. I learned a lot by just reading the code examples. Small tweaks to names and abstractions add up. No one captures that better than Kent.
PE funds optimize for a 4-7 flip and typically resort to aggressive cost-cutting and near-term optimization at the cost of the "soul" of the business or what made it successful in the first place.
Since I mostly work on Laravel projects or packages, I usually enable the Laravel Idea plugin. It's a paid plugin, but it's definitely worth the money since it can provide stuff like auto-completions for route names, request fields and more.
I think this is kind of leads to another way that Rust really helps with reliability, which is that if you're refactoring, I think Rust is really good at telling you all the places you need to update. I've done this sometimes where I would refactor something. I change the code. I change the return type or whatever it is, and then I just fix the compiler errors and until the compiler stops shouting. And then once I've done that, I've updated every place I need to update.
Immense scale, soulless optimization, and an insatiable thirst for growth dominate its behavior and discourse, leaving little room for the spirit and principles embodied by Steve Jobs and Steve Wozniak.
I- i- like, we are in a stage where I'm not building the code base to be perfect for me, but I wanna build a code base that is very easy for an agent to navigate.
But Amap is great! Especially since it can be used in English and I think it transliterates your searches to Chinese and the Chinese place names back to English
And so I would say, yes, we are just giving proxy names to things we don’t understand, but to dismiss that as some kind of, “Oh, they just don’t know…” It is actually quite the opposite.
I'm hopeful this framework will turn out to be a much more robust style for writing performance-sensitive code, especially over time and as compilers evolve.
But what we actually found was that none of those are actually hard. The whole idea of hard steps, that there are hard steps, is actually suspect. What's amazing about this model is it shows how important it is to actually work with people who are in the field.
Archeology, in this regard, is the worst enemy of this. So we put these names on cultures, we talk about how they evolve from one to another, we draw these lines where there aren’t any.
For the most part, it’s like these names are just utterly arbitrary, so you have no thing to latch on to. It’s not really a thing that our brain does very well to learn meaningless, arbitrary stuff. So what you need to do is build connections somehow, visualize a connection, and sometimes it’s obvious or sometimes it’s not.
The drone-like architecture of the multicopters is simple, but as the rotors are responsible for keeping the unit at a constant altitude with hover power, they represent an energy-intensive version of VTOLs. Their range is, therefore, short, typically less than 20 miles.
The best and easiest-found-by-optimization algorithms for solving problems we want an AI to solve, readily generalize to problems we’d rather the AI not solve; you can’t build a system that only has the capability to drive red cars and not blue cars, because all red-car-driving algorithms generalize to the capability to drive blue cars.
In traditional classroom settings, the development of skills often gets overlooked in favor of facts. Facts in the form of conjugation tables, word lists, grammar explanations, the names of verb tenses, and so on. All of these things are related to language, but it’s important to realize that knowing these things has nothing to do with speaking and using a language productively.
Other software I use to build stuff includes Netlify for hosting, GitHub for version control and collaboration, SVGO and TinyPNG for optimization, Kap for screen recording, GoatCounter for analytics, and Chrome and Firefox dev tools.
Other software I use to build stuff includes Netlify for hosting, GitHub for version control and collaboration, SVGO and TinyPNG for optimization, Kap for screen recording, GoatCounter for analytics, and Chrome and Firefox dev tools.
Other software I use to build stuff includes Netlify for hosting, GitHub for version control and collaboration, SVGO and TinyPNG for optimization, Kap for screen recording, GoatCounter for analytics, and Chrome and Firefox dev tools.
These are just three of the names on the list of thousands of Black men and women who lost their lives to the systemic racism that is woven into the fabric of this country.
I register all of my domain names through Hover. I think it's a little more expensive than some of the other options, but the UI is simple and reliable and not loaded with dark patterns like some of the cheaper competitors.
To me, this book is an illustration of the power of names. Today, in the era of Google, if you know the name of something, you can find out about it with a simple search. But if you don't know of what you're looking for, it suddenly becomes much harder to find it. Having in the back of your head the names of common algorithms that help you solve problems is really powerful.
We introduce Adam, an algorithm for first-order gradient-based optimization of stochastic objective functions, based on adaptive estimates of lower-order moments.
We introduce Adam, an algorithm for first-order gradient-based optimization of stochastic objective functions, based on adaptive estimates of lower-order moments.
I am a strong proponent of wanton notetaking - words, names, scraps, passages, whole articles, just dump it all in there - and I’d recommend this setup to anyone.
A korrent is a belief a person has stated in their own words: one
sentence stating the claim, backed by a quote and a source, kept at
korrents.com.
Under a name here, the quoted block is what they actually said.
The korrent beneath it is the claim those words support, in
korrents' wording — tap it to see the record, its source, and who
else holds it.
Nobody here wrote their own korrents. They are compiled from public
statements, and a person can change their mind, which is recorded too.
About the English under a post
Some people here publish in a language other than English. Where they
do, this site shows a machine translation beneath the post, in
this typeface — the site's own, not theirs.
The post itself is never changed, moved or hidden: what is set in the
serif above is exactly what the person published, and it is what to
quote them on. A translation can be wrong in ways that matter,
especially about tone.
Only the post's own words are translated. A quoted post, a linked
article and a belief on korrents.com
are left in their original language.