kingstnap
However, this doesn't mean its not magical when someone shows a next token predictor hand tune the weights in Qwen to make it generalize on math.
This is legit the stuff of science fiction if you presented it to mechanistic interpretability people a year or so ago.
But for real this is insane. Okay I need to get on this stat.
https://huggingface.co/burner22/numbat/discussions/13
This comment even has the receipts.
Idk how his AI figured this out, but the matrixes are literally just random data and the seed is the tensor name.
My own agent reproduced the findings.
Building something new is challenging, with or without AI its orders of magnitude slower than copying.
Building something new is also risky. Copying something with proven vitality is less so.
Given this almost tautologically you should expect way more clones than novel things.
This is the current one.
I haven't done french since high school. So totally chatgpt and google translate reading this to me, but I really don't get what VideoLAN has to do with this.
The law is crazy with a €55/TB tax on ingress to France (including local cdns). But it's got nothing to do with your video player.
> A tax is established based on the flows of digital data emitted by content providers and paid by the latter
> When a data emitter uses a content delivery network to cache content near end users, the amount of data emitted is considered equal to the sum of the outgoing flows from the delivery network
Imagine paying €8 to download Balder's gate 3 lol.
The 512 GB ultra seems too expensive. You can buy multiples of the cheaper one. Like 4 of the 128 GB max, or 2 of the 256 GB ultras for less. (And end up with more total bandwidth and storage and compute).
Unless there is some sort of restriction on number per customer.
It has better tuning towards not adding unhelpful caveats and irrelevant details. When I told it I was unhappy with some slop being dumped into some repo markdown files from Astra and Sol it actually went ahead and tastefully cleaned them up in line with how I expected them to be. Reductions are normally so bad from agents.
One big thing is that it really feels like talking to someone smarter than me. Other agents just seem extremely knowledgeable but Argon felt smart.
The outside implication for what is going on is that OpenAI has very clearly demonstrated insane capability in coming up with clever algorithms. You cannot deny some real level of super human like abstract reasoning here.
K-sat is the hard theoretical computer science problem. Nothing else is close in terms of how much it has been studied. There are journals and conferences for SAT.
It is really hard to come up with faster ways to solve it. So much so that people were convinced that you really can't go faster than exponential time. OpenAI has just shown a breakthrough construction better then decades of research by elite specialists was able to come up with.
Cross platform interpreter.
NumPy and SciPy really set the stage.
If you plan to look at moving things, like scrolling, which is a central part of using the device, 120fps makes the scrolling less blurry and therefore more pleasant to look at.
This is like asking why people have a hardon for cars with nicer materials for seats and interiors if a normal cheap plastic interior could pass all the same crash tests.
I'm not uninformed on this. I actually wasted a lot of time on it since my interests intersect at this. I've watched the numberphile interviews, read the drama, various takes etc.
What I'm saying is that its highly unlikely that in this case that Sébastien Bubeck from OpenAi decided to steal some critical idea from Buckmasters chat and feed it "Bell" the unreleased model that produced the solution.
> Obviously they need guidance to not get stuck in a dead end or have to brute force everything
Dude they did brute force everything. This is what I keep trying to say. They said they spent 130 billion output tokens.
This is quite literally like 500,000 books worth of text.
[0] https://help.openai.com/en/articles/6783457-what-is-chatgpt
If its memory bandwidth bound you need to raise your batch sizes to leverage more parallelism.
My point was explicitly that this didn't happen. I said that they took every possible path and published the one that worked. It's exactly throw shit at the wall and see what sticks.
> exactly after one of their users took it.
All parties agree that what happened is that someone had told someone working at OpenAI that Anthropic employees are working on Navier stokes. This is where the timing comes from.
---
What you are implying is this: OpenAI didn't do the throw shit at the wall approach as outlined in their release blog they actually secretely monitor everyones chat logs and found the idea in Tristan's Chat history and stole and published it.
Or some variant of this. You are only implying what you think is the truth, please state it outright. Don't beat around the bush.
There are 10k agents and Tristan's approach wasn't even groundbreaking novelty, it was literally based on existing published literature from Diego Córdoba and Luis Martínez-Zoroa that the agents have seen.
It's not just probable that agents would persue his particular niche of trying to construct a blowup. It's guaranteed if you just consider the basic setup.
When it comes to hard drugs I can understand why there are reasonable arguments for legalization of personal use.
Idk how many reasonable arguments extend the overton window to thinking that meth should be advertised seemingly all over the place. Because gambling ads are actually everywhere.
In principle AI Agents are superhuman at codeforces but tons of these vibe coded apps have ridiculously stupid performance problems.
The problem is a utilization of expertise thing, it can tell you with incredible detail about pros and cons and caveats to a problem if you ask about it. Truly superhumanly intelligent.
Calibration when it comes to actual decision making takes an incredible reduction in capability after that. Maybe it has something to do with the sycophancy or something but its like someone with exactly zero opinions (and therefore, taste).
My experience is that you can brainstorm with it on "Do A or B?" and it will produce a coinflip answer A, then if you continue and ask about a specific concern "what about this aspect of B?", it immediately 180s and flips back to B being way better than A.
Acting on this expertise when decision making points arise naturally for itself is where everything really nosedives. It's just gonna do whatever.
So when you are doing this "financial advice please" questioning, it is entirely hitting these weaknesses squarely in the face.
It really didn't come up with anything crazy. It was basically cutting spending and raising and creating new taxes. With the plan being to spread the pain based on tolerability and medium to long term sustainability. With actions like a fedral consumption tax with planned rebates depending on household wealth.
Its also tried to be careful around things that might create long term problems like dramatic education cuts and what not.
The reforms it gives are altogether pretty numerically radical but its not radical in the internet way where its a polarized solution. It's not a reddit style eat the rich or a business trickle down wealth solution.
I mean you can download and use verilator, openroad, yosys, ngspice, bsim-cmg, klayout and all this fun stuff.
The big blocker is really pdks. Where are you gonna get something that actually fabs? ASAP7 is just predictive.
And foundries are incredibly geopolitically important so they aint revealing anything.
If the context can be above average quality the result should be too as long as the AI is sufficiently smart. Especially if people iterate on the produced work to push it up in quality.
However if you don't have a good topic and don't iterate to improve the result then you get incredibly poor writing.
They tend to be wordy, long, and leak chatbot history or project management in unimportant ways that just distract.
I could point this out in my own code, but I don't need to since a vibe coded GitHub repo is linked in this very article.
> https://github.com/WordPress/openstation/blob/dd21eb131a2029...
// `const` now: the entry is no longer reassigned here, because
// `ensureGameRender()` moved inside the render callback below and
// its upgraded copy is consumed there.
* Phase 5 (boot decomposition) will move the runtime facade
* assembly here as well — a single `attachPublicApi(services)`
* function that wires every `wp.os.*` method in one place.
* Until then, the runtime assembly continues to live in
* `src/desktop.ts`.
* Phase 3 only ships drag-within-folder. Cross-folder drags
* (drop into a folder, drop out to root) and the Recycle Bin
* drop integration land in later phases.
// v1: no-op. Phase 4 will populate this from the push payload.
// We claim the event so a hosting environment that signs us up
// for a push subscription before the v2 PR lands doesn't see
// silently-dropped pushes — the empty handler still satisfies
// `Notification` capability checks on browsers that gate them.
Flash lite is very noticeably more idiotic than 3.8 Flash. It's why the quick AI overview is often much worse than the AI mode one.
Like in my experience the real burn is if there is some sort of feedback edge where an agent is producing stuff that eventually it has to reconsume.
Like if there is a dag of agents doing something its almost fine.
But if there is a loop somewhere without a huge amount of damping then you have an issue.
If you have ever been a human debugging in the middle of an agent it produces like so many commands and tests for you to do. Like way more than a human would.
I'm pretty sure they do the same thing to each other. Like the loop will amplify the yapping.
Regular chat, even on extra high thinking mode, is sort of unlimited. Idk at what point you hit the abuse gaurdrail but its really high whatever it is.
I have explicit instructions to tone it down. Less taglines, eyebrow text, subheadings, decorative spacing, pills, cards.
Hopefully this doesn't bleed into the chat...
> To keep the experience safe and welcoming, every published game undergoes safety screenings aligned with our Community Guidelines, along with user reporting.
Kind of cringe. I mean I suppose its one thing to stop people making CSAM but safety screens make it sound like you can't have any sort of mature game? Gore, violence, drugs, suggestive themes or nudity etc all disallowed or something? Everyone has to make a T for Teen game?
Like there is somehow redundancy in a fourier transform that makes it sub Linearithmic?
Which low and behold ->
130. Fourier transforms below n log n.