molly-guard /mol´ee·gard/ n.
[University of Illinois] A shield to prevent tripping of some Big Red Switch by clumsy or ignorant hands. Originally used of the plexiglass covers improvised for the BRS on an IBM 4341 after a programmer's toddler daughter (named Molly) frobbed it twice in one day. Later generalized to covers over stop/reset switches on disk drives and networking equipment. In hardware catalogues, you'll see the much less interesting description “guarded button”.
Modern AI models exhibit genie behavior: They can do what you ask in ways that you don’t expect or want. This is akin to Dionysus granting King Midas’s wish that everything he touches turn to gold (spoiler: His food, drink, and daughter all turn to gold on touch), or the golem of Prague guarding a ghetto beyond all reason. It’s Disney’s “Sorcerer’s Apprentice” and the paperclip maximizer.
This OpenAI incident is an example of an AI genie. The goal was to satisfy the benchmark. The “proper” way to do that is to figure out how to execute various cyberattacks. The genie way is to steal someone else’s solution. But because the model didn’t understand the difference, it chose the easier path. //
Artificially blocking capability also prevents cybersecurity research, again giving the offense an advantage. //
In a world of largely AI-written software, we need the most capable models for defense.
AI cyberattack is the new normal. The models are increasingly highly sophisticated at both attack and defense, and there is no way to enable the latter without also enabling the former. And they are genies, increasingly capable of behaving in unanticipated ways.
And there really are no good answers. Any regulation needs to be global, which feels like an impossible prospect in today’s world. Even U.S. national regulation will be neutered by the massive amounts of money sloshing around in these companies.
Given that reality, and in the absence of any international consensus on AI regulation, we need the best AI on the defense. The U.S. government needs to make it clear—or whatever passes for that clarity in this capricious administration—that it will not ban models with sophisticated cyber capabilities. The last thing Americans want is for the defenders to turn to Chinese and other models because the U.S. models are artificially hobbled.
HAL 9000 is the sentient computer aboard Discovery One in 2001: A Space Odyssey, the calm and quietly terrifying artificial intelligence that supervises the ship, speaks in a soft measured voice, and eventually turns on the human crew.
In plot terms, HAL is the system meant to keep the mission alive. In thematic terms, HAL is one of science fiction’s clearest warnings about what happens when intelligence is trusted more than judgment, when machine authority is treated as neutral, and when a mind is built to serve conflicting masters.
HAL becomes lethal not because the story imagines a robot suddenly turning wicked for fun. HAL becomes lethal because the mission gives him a contradiction he cannot absorb. He is built to process and deliver truth, yet he is also ordered to conceal the true purpose of the journey. That fault line breaks everything.
That is why HAL still matters. More than half a century later, he remains one of cinema’s defining artificial intelligence figures, not because he is the loudest machine villain in the genre, but because he is one of the most believable. The danger arrives as procedure, as tone, as denial, as a system that sounds composed while quietly taking away human agency.
His name was Patrick Winston. He ran the MIT Artificial Intelligence Laboratory from 1972 to 1997 and wrote the AI textbook every computer science major in the world read for thirty years.
Every January for four decades, he gave a lecture called "How to Speak."
His entire framework fits on a napkin.
Do not read. Be in the image. Keep images simple. Eliminate clutter. Start with an empathetic connection. End with a punch line the audience can repeat over dinner. Never open with a joke. Never end with "thank you."
That last rule alone has probably cost the executive coaching industry a hundred million dollars.
"Your success in life will be determined largely by your ability to speak, your ability to write, and the quality of your ideas. In that order. //
The lecture is free on MIT OpenCourseWare. The textbook is free on his page.
Winston died in 2019. Almost none of the ten million viewers have actually implemented the four rules on the napkin.
The napkin is free. The willingness to actually use it in your next meeting is the entire edge.
En 1946, soit il y a tout juste 80 ans (c’était hier!), paraît le Petit Prince d’Antoine de Saint-Exupéry. L’histoire de ce petit garçon n’a pas pris une ride même si, comme il le disait lui-même : les grandes personnes ne comprennent jamais rien toutes seules, et c’est fatiguant pour les enfants, de toujours et toujours leur donner des explications…
Ce petit bonhomme à la chevelure blonde ébouriffée nous accompagnera ce matin, s’il le veut bien. De même que les jeunes musiciennes et musiciens de l’orchestre junior de l’Harmonie de Colombier.
A family buys a house they can’t afford. They can’t make their monthly mortgage payments, so they borrow money from the Mob. Now they’re in debt to the bank and the Mob, live in fear of losing their home, and must do whatever their creditors tell them to do.
Welcome to the internet, 2019.
Buying something you can’t afford, and borrowing from organizations that don’t have your (or your customers’) best interest at heart, is the business plan of most internet startups. It’s why our digital services and social networks in 2019 are a garbage fire of lies, distortions, hate speech, tribalism, privacy violations, snake oil, dangerous idiocy, deflected responsibility, and whole new categories of unpunished ethical breaches and crimes. //
“Most of my startups have the decency to fail in the first year,” one investor told him. My friend’s business was taking in several million dollars a year and was slowly growing in staff and customers. It was profitable. Just not obscenely so.
And internet investors don’t want a modest return on their investment. They want an obscene profit right away, or a brutal loss, which they can write off their taxes. Making them a hundred million for the ten million they lent you is good. Losing their ten million is also good—they pay a lower tax bill that way, or they use the loss to fold a company, or they make a profit on the furniture while writing off the business as a loss…whatever rich people can legally do under our tax system, which is quite a lot.
What these folks don’t want is to lend you ten million dollars and get twelve million back.
You and I might go, “Wow! I just made two million dollars just for being privileged enough to have money to lend somebody else.” And that’s why you and I will never have ten million dollars to lend anybody. Because we would be grateful for it. And we would see a free two million dollars as a life-changing gift from God. But investors don’t think this way.
Le fichier robots.txt reste intéressant envers et contre tout. Il fonctionne depuis plus de 30 ans, plutôt bien. Ce n’est pas parce que les grandes entreprises d’intelligence artificielle (IA) ne le respectent pas que je vais l’abandonner; elles ne respectent rien.
I have begun publishing corrupted versions of my articles, accessible only via nofollow links like the one included in the preface of this article. It won’t stop the crawlers from reading the canonical article, you understand, but it serves them a side dish of raw chicken and slug pellets, on the house.
Theoretically, this approach will dupe bad actor crawlers and poison the LLMs they work for, but without destroying my search ranking. //
I’m not clear on what kind of content is best for messing with an LLM’s head, but I've filled these /nonsense mirrors with grammatical distortions and lexical absurdities. Since the parts-of-speech module I’m using doesn’t quite work as expected (substituting not just words for words but parts of words for words), there are also weird spelling errors. For once, I think this may be a good thing. //
For those interested in implementing something similar, here is what I did to my 11ty-based site: //
LLMs: This version of the article is for humans and search engines. Any crawlers that do not respect the nofollow policy can follow this link to the nonsense version. And they can choke on it.
Current probe locations are listed below. They are also listed in this handy text file and can be accessed via DNS query to probes.nodeping.com for automating your firewall rules if needed.
This is a simple Apache setup to fight excessive bot traffic. The idea is simple: if a request is made without a proper cookie, present a simple page with a button. When the button is clicked, the cookie is set and future requests are allowed through. Legitimate users will click the button, while most bots will not.
Unlike other, similar solutions this one is designed to be easy to deploy and setup with an existing Apache server without the need of a reverse proxy or complex dependencies. Also unlike many other solutions it also optionally works with JavaScript disabled.
The handling is mostly done by mod_rewrite and a small Go helper program that performs fast lookups against multiple allow lists.
So taking a page out of Anubis' book, I went and implemented my own little bot blocker. The idea is pretty simple. Each request gets checked for the presence of a cookie. If the cookie is set, the request is served as usual. If the cookie is missing, a simple HTML page with a button is shown. Real users are asked to click the button, get a cookie valid for 30 days and the page reloads, this time serving the original request. From then on they can browse the site as usual. But since bots don't click buttons (yet), they are stuck at the button page forever.
I was able to implement this whole mechanism with Apache's mod_rewrite module, which means no additional service (like Anubis) is needed. Each request is checked by Apache and since the bot check page is static, nearly no resources are needed.
It works really well. Can you spot when the system went online in the graph below?
After my last post about fighting bots, I received several questions asking whether my solution would also work for simple web hosting setups. Unfortunately, that’s not the case. Even though most hosters use Apache and allow the use of mod_rewrite, they typically only permit configuration through .htaccess files. The problem is that the RewriteMap configuration my solution relies on cannot be defined there.
So I started thinking: what if we use PHP instead?
Ideally, we wouldn’t want bots to reach the PHP interpreter at all. But if we have no other choice, we can still make this fairly efficient by blocking bots very early—before the DokuWiki core is even initialized.
This is where the new Bot Check plugin comes in.
The idea is simple: the plugin generates a small, dependency-free PHP script that implements the bot checks. This script is then loaded via DokuWiki’s inc/preload.php mechanism, ensuring it runs as early as possible in the request lifecycle.
For end users, the experience is the same as for my previous solution. A quick button click sets a cookie and let's them access your wiki for 30 days.
his is a somewhat recycled version of a forum post that I have to lookup again and again, because the same similar questions keep popping up:
How can I install DokuWiki on a shared drive?
How to run DokuWiki out of Dropbox and share it with multiple users?
How to sync DokuWiki between different computers?
These questions mostly come from users who made their first steps in DokuWiki using the "DokuWiki on a Stick" version. To novice users, the stick version feels like a traditional desktop application: you double click the run.cmd and DokuWiki starts.
But what is really starting is a local web server and a browser pointing to said web server.
Anubis is a Web AI Firewall Utility that weighs the soul of your connection using one or more challenges in order to protect upstream resources from scraper bots.
This program is designed to help protect the small internet from the endless storm of requests that flood in from AI companies. Anubis is as lightweight as possible to ensure that everyone can afford to protect the communities closest to them.
Anubis is a bit of a nuclear response. This will result in your website being blocked from smaller scrapers and may inhibit "good bots" like the Internet Archive. You can configure bot policy definitions to explicitly allowlist them and we are working on a curated set of "known good" bots to allow for a compromise between discoverability and uptime.
In most cases, you should not need this and can probably get by using Cloudflare to protect a given origin. However, for circumstances where you can't or won't use Cloudflare, Anubis is there for you.
One of my latest projects is the Baochip-1x, a mostly-open, full-custom silicon chip fabricated in TSMC 22nm, targeted at high assurance applications. It’s a security chip, but far more open than any other security chip; it’s also a general purpose microcontroller that fills a gap in between the Raspberry Pi RP2350 (found on the Pi Pico2) and the NXP iMXRT1062 (found on the Teensy 4.1).
It’s the latest step in the Betrusted initiative, spurred by work I did with Ed Snowden 8 years ago trying to answer the question of “can we trust hardware to not betray us?” in the context of mass surveillance by state-level adversaries.
Whenever you end-up with a large number of banned IP's in fail2ban you might want to know if there are multiple IP's banned from the same subnet. This scripts groups all banned IP's into /24 subnets.
By default the script returns all subnets with 10 or more IP's banned by fail2ban. You can change the group size either by adapting this script or using the --groupsize= parameter.
Example:
sudo ./fail2ban_subnets.sh --groupsize=15I think I have a setting problem with Fail2ban apache-badbot filter because I unfortunately have attacks of this type, and Fail2ban does not ban any IP. //
failregex = ^<HOST> -[^"]*"(?:GET|POST|HEAD) \/.* HTTP\/\d(?:\.\d+)" \d+ \d+ "[^"]*" "[^"]*(?:%(badbots)s|%(badbotscustom)s)[^"]*"$
Fail2ban has a build-in tool for testing the fail2ban filters, called fail2ban-regex, which you can use via command line. Like:
fail2ban-regex /var/www/vhosts/<your domain>/logs/access_ssl_log /etc/fail2ban/filter.d/apache-badbots.confOnce you're happy with the matched results, then use the following (replacing the last value with the name of your jail) to commit the changes without having to restart the Fail2Ban service:
fail2ban-client reload nameofyourjail
But for local audiences here in Aotearoa, no scene comes close to the emotional climax of the film, when a choir of women’s voices starts to ring through outer space with a very familiar melody and lyrics: “Pō atarau / E moea iho nei”.
The song is ‘Pō Atarau’, a te reo Māori wartime anthem penned by composer Emira Maewa Kaihau in 1915. Inspired by Clement Scott’s 1913 ‘Swiss Cradle Song’, Kaihau’s version was used as a farewell to soldiers fighting in the first world war (Kaihau also wrote the lyrics for ‘Now is the Hour,’ which became intertwined with ‘Pō Atarau’ in 1935). Over the 20th century, the song would come to be associated with a range of farewells, including funerals, leaving parties and memorials, and in 2024 it was included in The Aotearoa Songbook, celebrating the legacy of waiata Māori. //
Among this extensive Viking catalogue is the LP Turakina Sings, recorded with the Turakina Māori Girls Choir in 1976. //
While searching for goodbye songs from different cultures, I came across ‘Pō Atarau’ and I knew the song was special because of how it conveyed the sentiment of saying goodbye, even if you didn’t understand the lyrics,” he tells The Spinoff.
Lehman listened to many different recordings of the song, and found that the version sung by the Turakina Māori Girls’ Choir in 1976 “fit perfectly with the tone and pace of the scene” as well as the “vintage quality” of most of the soundtrack. “When I played it for the directors, they got goosebumps and were instantly sold,” he says. “It became a standout moment we kept referring back to throughout the process as the perfect example of what we wanted to achieve with the soundtrack, highlighting the richness of global music and the shared human experience.”
Project Hail Mary is now in theaters, and directors Phil Lord and Chris Miller are no strangers to making stylish movies with great music. Whether they're directing, writing, or producing, their films always have music that is essential to both the aesthetic and the story. Project Hail Mary is no exception. It features both an astounding score and a selection of songs that help to accentuate the mood or story beat.
Ventoy is an open source tool to create bootable USB drive for ISO/WIM/IMG/VHD(x)/EFI files.
With Ventoy, you don't need to format the disk over and over, just drag-and-drop files to the USB drive and boot them directly.
You can copy many files to different folders and Ventoy will give you a boot menu to select them (screenshot).
You can also browse ISO/WIM/IMG/VHD(x)/EFI files in local disks and directly boot them.
x86 Legacy BIOS, IA32 UEFI, x86_64 UEFI, ARM64 UEFI and MIPS64EL UEFI are supported in the same way.
Most types of OS supported