modeless

14 ms·

331.
by modeless 8mo ago
Oh cool, I didn't know they added their Street View equivalent to the web version. The animated transitions are much better than Google's.
332.
by modeless 8mo ago
Street View is such a missed opportunity. In 2007 it was visionary and essential to create the map data that allowed Google Maps to win. In 2026 it is a symbol of Google's stagnation. Essentially zero improvement in user experience for
333.
by modeless 8mo ago
It's clear that in the general case "predict the next word" requires arbitrarily good understanding of everything that can be described with language. That shouldn't be mysterious. What's mysterious is how a simple
334.
by modeless 8mo ago
> The models have a very large user base, but very narrow engagement and stickiness, and no network effect or any other winner-takes-all effect so far that provides a clear path to turning that user base into something broader and durabl
335.
by modeless 8mo ago
That's what it says when it doesn't like the look of your IP. I get it when I'm on my VPN.
336.
by modeless 8mo ago
They did the exact same thing with Mac Pro in 2019. I notice they don't say they'll stop manufacturing the Mac Mini anywhere else. This is a political thing and will change with the political winds.
337.
by modeless 8mo ago
Article is wrong. Tablets/pods are silly. The most important thing to improve dishwasher performance is to put detergent in the prewash compartment as well as the main compartment, which you can't do with pods. Second most importa
338.
by modeless 8mo ago
You don't need exclusive fullscreen on Windows to bypass the compositor. Fullscreen borderless windows also bypass the compositor. And in newer Windows versions the compositor can be bypassed even in regular windows using hardware over
339.
by modeless 8mo ago
> isn't X – it's Y What is this AI slop doing at the top of HN? Come on, you don't even have to click through to know it's slop! It even has an en dash right in the title!
340.
by modeless 8mo ago
Revisiting this and I have to say the sheer number of demos on this page is incredible. I love this one in particular as I remember this classic scene from Veach's thesis: https://erichlof.github.io/THREE.js-PathTracing
341.
by modeless 8mo ago
This is neat. In the demos I would suggest making mouse/finger drag orbit the camera around the scene instead of panning. Panning can be done by a 2D image transformation so it doesn't show off the 3D nature of the renderer.
342.
by modeless 8mo ago
Awesome project! What did you do afterwards?
343.
by modeless 8mo ago
He has been wrong about timelines and about what specific approaches would ultimately solve ARC-AGI 1 and 2. But he is hardly alone in that. I also won't argue if you call him smug. But he was right about a lot of things, including mos
344.
by modeless 8mo ago
Here is what the original paper for ARC-AGI-1 said in 2019: > Our definition, formal framework, and evaluation guidelines, which do not capture all facets of intelligence, were developed to be actionable, explanatory, and quantifiable, r
345.
by modeless 8mo ago
Now do one where you have to withdraw your card from the machine before it starts beeping obnoxiously at you but the screen keeps trying to trick you into withdrawing too early.
346.
by modeless 8mo ago
Hiive and Forge Global are the ones I know of. You must be an "accredited investor" which means nothing at all except that you have a million dollars or make $200k/yr.
347.
by modeless 8mo ago
Worth keeping in mind that in this case the test takers were random members of the general public. The score of e.g. people with bachelor's degrees in science and engineering would be significantly higher.
348.
by modeless 8mo ago
$14B revenue run rate is the interesting number here.
349.
by modeless 8mo ago
It's still useful as a benchmark of cost/efficiency.
350.
by modeless 8mo ago
Why are they obscuring the price? It must be outrageously expensive.
351.
by modeless 8mo ago
François Chollet, creator of ARC-AGI, has consistently said that solving the benchmark does not mean we have AGI. It has always been meant as a stepping stone to encourage progress in the correct direction rather than as an indicator of rea
352.
by modeless 8mo ago
Exactly. I think there's an opportunity right now to reinvent public forum moderation with LLMs. Karpathy's recent post is going in the right direction I think: https://karpathy.bearblog.dev/auto-grade-hn/
353.
by modeless 8mo ago
I have also been wanting this. Thinking about building something. Of course it takes more than code, it needs a community, and I don't know how to bootstrap that. But it seems to me there ought to be a lot of people dissatisfied with t
354.
by modeless 8mo ago
For this purpose Google offers "Inactive Account Manager" AKA a dead man's switch.
355.
by modeless 8mo ago
I disagree. A year ago the models would not come close to doing this, no matter what tools you gave them or how many tokens you generated. Even three months ago. Effectively using tools to complete long tasks required huge improvements in t
356.
by modeless 8mo ago
It means that it's not directly copying existing C compiler code which is overwhelmingly not written in Rust. Even if your argument is that it is plagiarizing C code and doing a direct translation to Rust, that's a pretty interest
357.
by modeless 8mo ago
They are doing that too. https://red.anthropic.com/2026/zero-days/
358.
by modeless 8mo ago
There seem to still be a lot of people who look at results like this and evaluate them purely based on the current state. I don't know how you can look at this and not realize that it represents a huge improvement over just a few month
359.
We tasked Opus 4.6 using agent teams to build a C Compiler(anthropic.com)
735 points by modeless 8mo ago|738 comments
360.
by modeless 8mo ago
I also wasn't that familiar with it, but the Opus 4.6 announcement leaned pretty heavily on the TerminalBench 2.0 score to quantify how much of an improvement it was for coding, so it looks pretty bad for Anthropic that OpenAI beat the

More ›