The threat feeds we offer are in alignment with what the user expects and intends. Mandatory restriction of DNS resolution by a third party (not the user, not Quad9) is expressly what users do NOT want, regardless of the outcome. Once such restrictions are enacted and treated as normal, the channel for making demands will fill with a flood of "objectionable" names from every corner for every reason. Quad9 has already received demands from the French regulator to block sites which deliver adult content found to be objectionable to minors. We see an extremely slippery slope which has already moved out of piracy as the scoped intent which is this case in particular.

More to the point: prohibiting content via the DNS is a useless exercise when it is done out of alignment with the end users' goals - they will simply use some other recursive resolver (or their own) probably without the benefits of being oriented towards security and privacy. We have no monopoly on DNS resolution - there are thousands of others open to public use, and anyone can set up their own, for free, with no mandatory requirements by governments or regulators. This case (and any others like it) only drives out law-abiding organizations who are trying to promote better cyber hygiene and who want to keep the DNS as a single, un-splintered high-integrity pointer system.

An editing pass to make the prose less "punchy" would be great.

The call to action says "Copy prompt", not "Unlock prompt", "Sign in to copy", or "Get Creative to access prompts". I didn't even go past the signup to see if the prompts themselves are actually free and work without a subscription to your hosted MCP server. There's a GitHub link in the header suggesting an OSS tier, at the least. That repo may include "prompts" folders, but no prompts are there. The README doesn't even hint that this is a paid library of assets/prompts until the very end, with the misspelled "enhanched" sentence under the "Pro" header. The only prompts that appear to be in the repo (other than those that configure an agent how to use the paid and hosted MCP) are basic design guidelines.

Getting back to the website: your hero text directly invites users to "Try each example and copy its implementation prompt." Which immediately demands the aforementioned signup flow that I didn't bother completing.

All of that just turned me right off.

Your examples are nice, if a bit generic. But if you intended this to be a loss-leader for the more "enhanched" UI/UX skills, it isn't doing that job.

I think we need to keep pushing back against this idea that "now it's the AI coding era, everything gets coded with AI." Let's keep building things, especially passion projects carefully by hand and normalizing this. Lots of us are still coding without AI.

It would be like me practicing surgery 1 hour a week and saying "well surgery only has a very small place in my 80-hour work week" in response to flagrant failure of medical care.

Fundamentally, I know that HN has plenty of people who will react with "ha-ha, human authors should pack their bags", just like what we keep telling mathematicians, because we see the discipline as just another market that can be conquered by automation or another benchmark to track. But literature, similarly to most other forms of higher-brow entertainment, isn't just about reducing unit-cost or hacking engagement metrics. For the same reason why a genuine van Gogh painting is worth a lot more than a nearly-perfect replica, people who care about the craft usually care about it telling genuine human stories and celebrating the author as much as the result.

Now, the mass consumer may not care; if you have an infinite Harry Potter sequel machine, many will partake. But then, I guess the question is, if you build the sequel machine, are you leaving the world better off or not? Do you want to be interacting exclusively with machines down the line? And if there's no inherent value to humanity, why should the machines be compelled to interact with you?

In the 1970s, you began to see the emergence of the drummer "playing to the click" and the main motivation behind this was that some number of the rest of the performers were not in the room and/or not recording at the same time as the drummer. If you let the drummer just do their human thing, the timing will wander a bit in ways that are hard to follow and respond to if you were not in the room. So producers and engineers began to use click tracks to ensure that now matter when/where the differ performers actually recorded their work, everyone would be locked together.

If you record all together in a single space at a single time, you generally don't need this (unless the musicians are not very good and the drummer still wanders a lot).

So it's a phenomenon that didn't begin with DAWs and computers - that technology just took the overall approach even further.

https://www.switch-bot.com/products/switchbot-bot

100% squashing gambling has generally not been successful, but the strategy of fencing it away from most of society and keeping it in a set of defined places that require people to actively seek it out has been successful enough that people forget why we don't allow gambling in every convenience store in the country. I'd say the additional "convenience" of having it in your pocket all the time is a new nuance, but doesn't necessarily change the fundamental equations.

We will relearn. Our grandchildren will forget again. (Unless things have really changed.)

Essentially, the PHP team is deflecting and abdicating their responsibilities in all this mess they created. "Chainguard broke embargo", "support us with money", "we are just volunteers", "we had knowledge years ago but no POC because we wanted to ship features and not security fixes". All just deflections.

This fiasco shows me that I would never use PHP for any project I'm part of. The PHP team royally messed up, and they are deflecting the blame left, right, up, and down.

People seem to forget that LLMs copied human writing. They also tended to copy people with more proper or formally trained writing styles. There are so many journalists and writers that talk about how they've had to 'stupify' how they write so they don't get accused of being an LLM.

It's like we're running idiocracy on fast forward.

To be honest, I'd be fine with some AI-assisted writing and might even be interested in reading one of those. The scientific sources seem to be researched well, not hallucinated. But asking the author "Alexander Crespo" under a reddit post, to what extent they used AI in writing, results in them blocking you. Makes me think it is more AI-generated than they wanna admit.

We built a way to sample diverse ideas in machine learning - ideas that make scientific sense, but that no existing research community is positioned to propose.

In short, the way it works is like this:

1. Take 16,068 LLM papers (NeurIPS, ICLR, ICML, ACL, EMNLP, 2017 to 2025) are broken into 273 "idea atoms". Every paper, and every candidate idea, becomes a set of atoms.

2. Train a Coherence Model: a 42.8M-parameter transformer trained on papers as sequences of atoms. Basically a language model whose tokens are ideas.

3. Train a Cognitive-Availability Model: a dual encoder trained on the publication histories of 39,791 researchers, which predicts where each person could plausibly go next.

4. Use those models to score sampled tuples using reciprocal rank fusion (you want ideas that go together well, but that are not cognitively available to anyone).

We tested it in a bunch of ways. One interesting one was that we took all the papers at the end of 2024, then checked which high-scoring tuples showed up in 2025 papers. It worked well - one examples is an ICLR 2025 oral, goes from rank 1,489 with coherence alone to the top 25 of 1.14M samples.

A funny one: if you ask Claude to come up with an ML research idea, it is often about interpretability. If you then run an autoresearch pipeline with some A100s on the idea, the result tends to be interpretability 8/10 times. Our Alien Sampling method stayed at 0 interpretability ideas even after letting claude cook on them.

We also paid 50 people with graduate degrees in CS/AI to judge the research reports and found that Alien’s reports had the best average quality rank and produced 3 of the top 5 reports.

Oh and then we scaled it way the hell up to a couple million papers and 10,000,000's of concept elements and generated 25,000 new research proposals that you can search over here: https://lacuna.tiptreesystems.com/explore?view=hypotheses

Code (MIT): https://github.com/alejandrohdez00/alien-space-of-science

LLMs generally cannot report the outputs of their own activation functions. Taking the output of softmax as if it were a probability score is considered a first-year PhD student misconception.

Others are LPEs, but two have been deleted by now, so can't be sure of the details.

TrendMicro's site opens up with a big big "Advancing AI-Native Security with Anthropic". Which checks out I guess.

I feel like there's enormous potential still uncovered in the coding agent space for such highly-discounted "spot" style capacity. Give me a way to have agents continuously crawl over my codebase looking for bugs, come up with test cases that prove the bug, open a PR. I don't care if it takes 12 hours to find something that would take 5 minutes if prompted with "on-demand" / "high-priority" requests, because I'm not prompting for it directly, I'm not aware of how much time it takes, all I see is continuous improvement.

> You should patch your servers anyway. There is a real vulnerability here, but exploiting it is significantly harder than people are making it out to be.

The take-home message is about human herding behavior being reinforced by LLM use:

> There's something important to be said here about how many people have Claude/GPT make design decisions for them, and how AI is the primary decision maker in the world right now. You've got this bizarre case of "Decision Schelling Points" where a method of solving a problem is extremely overrepresented, because it's what Claude does out of the box.

On a very similar theme, I recently built a “lossy git”, which stores your code by training a neural network, then it generates the code back to you on checkout

* web demo: https://nicobrenner.github.io/lossy-git/demo.html

* cli version: https://nicobrenner.github.io/lossy-git/

The web demo runs entirely on the browser, both the version control and the neural network training/inference. You can commit and checkout both images or code - you can also paste/“upload” your own images, no server requests (everything clears from the page on reload)

But I question its utility on other platforms, because a lot of the things that makes PowerShell cool on Windows doesn't actually work in Linux or macOS.

Eg: Get-Service (and associated cmdlets). Very handy cmdlet that I use all the time on Windows. On Linux, you'd expect that it'd work with systemd units, but no, that command doesn't exist on the Linux version. Same with Get-NetAdapter, or Get-NetTCPConnection, or Get-Disk and so many other handy cmdlets that I use all the time on Windows, are all missing.

Yet, some cmdlets do exist, like Get-Process or Get-Uptime etc, so I don't understand why many of the commonly used Windows cmdlets haven't been ported over.

Another annoying quirk is how PowerShell on *nix hands off some of the built-in aliases to their native versions. Like on Windows, I always use 'ls' because ain't nobody got time to type Get-ChildItem. Unfortunately when I run this on *nix PS, it runs the native version. And this is a problem because when I pipe 'ls' into a for loop or another cmdlet for further processing, it naturally, doesn't work. To make things worse, this native-or-alias decision making isn't even consistent. Like 'pwd' for example actually runs the PowerShell version, whereas 'echo' runs the native version.

Basically, as much as I love PowerShell on Windows, on *nix it is very annoying to work with. I would much rather work with native *nix shells and scripting languages. And if I really wanted the structured output of PS, I would run Nushell instead, which is far more consistent and useful than the half-baked product that *nix PowerShell is.

It’s not just Airbnb the company that moved away from it, it’s the Airbnb hosts that moved away from it as soon as opportunists started becoming mega landlords renting out ikea-furnished party houses. It hasn’t been the original “person with a spare room for a weekend” Airbnb in a long, long time.

Funny how the article has zero citations or notes.

So it goes.

The lack of transfer I don't think is out of unwillingness from the old guard, but rather due to the greenhorns throwing everything that came before their time out without questioning the reasoning behind any of it (Chesterton’s Fence) and assuming that whatever they came up with was superior and better fit for the modern era ("those old principles can't possibly hold, we have touch screens now!").

It's also been made worse by UI design roles having been taken by those with backgrounds in unrelated design disciplines, as well as sites like Dribbble which reward pretty mockups over working products. This leads to overemphasis on the visual aspect of the UI and not enough on its functional and usability aspects.

Our American obsession with jail is nuts. The violation at hand was “misleading state citizens about how it safeguards and collects user data.”

Lots of folks at Meta should be criminally investigated. But not for this, not unless someone willfully signed off on a lie.

> This will keep happening until those at the top are held personally responsible

Perfect is the enemy of the good.

That being said I don't particularly understand your conception of a method. They are both methods. The getter ought to have a shorter name since it is probably called more frequently and because it inherently has less of an impact -- this goes along with the principle that names that demand attention should be larger and easier to notice.