PLAY PODCASTS
Chaos Computer Club - recent events feed

Chaos Computer Club - recent events feed

4,903 episodes — Page 41 of 99

Von wegen Eisblumen! Wie man mit Code, Satelliten und Schiffsexpeditionen die bunte Welt des arktischen Phytoplanktons sichtbar macht (39c3)

Die Arktis ist eine Region, in der die Sonne monatelang weg ist, dickes Meereis den Weg versperrt und deshalb Forschungsdaten ziemlich rar sind. Kompliziert also, herauszufinden was im Wasser blüht! Mit einer Kombination aus Satellitenbildern, Expeditionen und Modellsimulationen auf Hochleistungsrechnern versuche ich, das Verborgene sichtbar zu machen: die faszinierende, farbenfrohe Welt des arktischen Phytoplanktons. Im Arktischen Ozean wird immer deutlicher, wie stark die globale Erwärmung den Rückgang des Meereises und das marine Ökosystem beeinflussen. Winzige Organismen, das Phytoplankton, bilden die Grundlage des Nahrungsnetzes durch den Aufbau von Biomasse und spielen so eine zentrale Rolle im globalen Kohlenstoffkreislauf. Dabei werden sie in der Arktis stark von den jahreszeitlichen Schwankungen der Polarnacht/-tag, der Meereisausdehnung und der sich verändernden Umwelt beeinflusst. Doch das Phytoplankton ist nicht nur ökologisch bedeutsam, sondern auch erstaunlich vielfältig und farbenfroh – wie eine bunte Blumenwiese im Ozean! Spannend bleiben dabei auch die Fragen, was die Vielfalt des Phytoplanktons ausmacht, wie diese eine Anpassung an die Umweltveränderungen ermöglicht und wie sich das arktische Ökosystem unter verschiedenen Klimawandelszenarien entwickeln könnte. Dieser Vortrag lädt euch ein, in die eisigen Welten des arktischen Ozeans einzutauchen, um dem grundlegenden Baustein des arktischen Ökosystems, dem Phytoplankton, auf den Grund zu gehen. Licensed to the public under http://creativecommons.org/licenses/by/4.0 about this event: https://events.ccc.de/congress/2025/hub/event/detail/von-wegen-eisblumen-wie-man-mit-code-satelliten-und-schiffsexpeditionen-die-bunte-welt-des-arktischen-phytoplanktons-sichtbar-macht

Dec 29, 202537 min

Learning from South Korean Telco Breaches (39c3)

2025 was a bad year for South Korean mobile network operators. All three operators (SK Telecom, KT, LG U+) were affected by breach in some part of their respective network: HSS of SK Telecom, femtocells of KT. Meanwhile, handling of the breach by each operators and post-mortem analysis of each breaches have stark differences. The technical details and implemented mitigations are often buried under the vague terms, and occasionally got lost in translation to English. In this talk, I will cover the technical aspects of SK Telecom and KT's breach, and how the operators are coping to the breach and what kind of measurements have been performed to secure their network. This talk will cover the public information and experiments related to the South Korean telco breaches in 2025. This talk will cover SK Telecom's HSS breach (final results announced), KT's femtocell breach (investigation ongoing) and related operator billing fraud, and revisit Phrack report on KT and LG U+ breach. We also give a light on the detail regarding the implemented mitigation and diaster response of each operators. SK Telecom's HSS breach is attributed to a variant of BPFDoor malware, resulting leakage of critical operator data related to subscriber authentication and accounting. They replaced the SIM cards of all 23 million subscribers, and implemented additional mechanism to track the possible cloning of the SIM card. We analyze the aftermath and how it will effectively protect against the said attack. KT's femtocell and operator billing breach (investigation still ongoing as the time of writing) is attributed to the mismanagement of KT's femtocell, allowing an external attacker to mimick the behavior of KT's legitimate femtocell and use as a cellular interception device. This is a modern implementation of the remarkable research "Weaponizing Femtocells" back in 2012, and new cellular technologies like VoLTE have changed the possible attack vectors. We provide a possible theory on how the attack would be possible, based on the publicly available information and previous researches. Finally, we also cover the characteristics of South Korean mobile market and how the media caused the inaccurate analysis and FUD (fear, uncertainty, and doubt). In particular, how SMS-based 2FA is tied to personal authentication and how everything is strongly bound to the personal identity. Early media reports could be attributed to the information "lost in translation" and inaccurate information in English-language articles when the details of the breach were not widely shared. We try to correct the information (also in the official incidence report) and showcase how not to report the breach in general. Licensed to the public under http://creativecommons.org/licenses/by/4.0 about this event: https://events.ccc.de/congress/2025/hub/event/detail/learning-from-south-korean-telco-breaches

Dec 29, 202549 min

Omnibus Halbgarer Machenschaften (OHM #23) (39c3)

Dec 29, 202546 min

Supplements und Social Media – wenn der Online-Hype zur realen Gesundheitsgefahr wird (39c3)

Dec 29, 202558 min

Making the Magic Leap past NVIDIA's secure bootchain and breaking some Tesla Autopilots along the way (39c3)

Dec 29, 202557 min

Programmierte Kriegsverbrechen? (39c3)

Dec 29, 20251h 0m

Learning from South Korean Telco Breaches (39c3)

Dec 29, 202549 min

Omnibus Halbgarer Machenschaften (OHM #23) (39c3)

erdgeist & monoxyd denken laut. Aufgrund des großen Erfolgs soll das jetzt auch beim Congress versucht werden. Themen? Ja! Wahrscheinlich irgendwas mit so... Dingen, die gerade passiert sind und zu denen mal was gesagt werden muss. Besser wir als Lanz & Precht! Es gibt außerdem einen besonderen Anlass: 23! (Wo kommt das eigentlich her?) Licensed to the public under http://creativecommons.org/licenses/by/4.0 about this event: https://events.ccc.de/congress/2025/hub/event/detail/omnibus-halbgarer-machenschaften-ohm-23

Dec 29, 202546 min

Programmierte Kriegsverbrechen? (39c3)

Die automatisierten Zielsysteme des israelischen Militärs zeigen gut, wie KI-baserte Kriegsautomatisierung aktuell aussieht, was daran falsch läuft und warum wir Techies uns einmischen müssen Das Thema „KI in der Militärtechnik“ und die Beziehung zwischen Mensch und Maschine ist seit Jahrzehnten ein Thema in der Friedensbewegung, der Konfliktforschung, der Philosophie, den Sozialwissenschaften und den kritischen Data & Algorithm Studies. Doch in den letzten Jahren wurden Waffensysteme mit KI-Komponenten entwickelt und auch praktisch in bewaffneten Konflikten eingesetzt. Dabei reicht die Anwendung von Drohnensteuerung über optische Zielerfassung bis hin zur logistischen Zielauswahl. Am Beispiel KI-gestützter Zielwahlsysteme, die vom israelischen Militär seit Mai 2021 und insbesondere jetzt im Genozid in Gaza eingesetzt werden, können die aktuellen technischen Entwicklungen aufgezeigt und analysiert werden. Im Fokus dieses Talks stehen vier KI-unterstützte Systeme: Das System Gospel zur militärischen Bewertung von Gebäuden, das System Lavender zur militärischen Bewertung von Personen, das System Where's Daddy? zur Zeitplanung von Angriffen und ein experimentelles System auf Basis großer Sprachmodelle zur Erkennung militärisch relevanter Nachrichten in palästinensischen Kommunikationsdaten. Auf Basis der Aussagen von Whistleblower:innen des israelischen Militärs und Angestellten beteiligter Unternehmen wie Amazon, Google oder Microsoft sowie internen Dokumenten, die durch investigative Recherchen von mehreren internationalen Teams von Journalist:innen veröffentlicht wurden, können die Systeme und Designentscheidungen technisch detailliert beschrieben, kritisch analysiert sowie die militärischen und gesellschaftlichen Implikationen herausgearbeitet und diskutiert werden. Dabei entstehen auch Fragen bezüglich Verantwortungsverlagerung durch KI, Umgehung des Völkerrechts sowie die grundsätzliche Rolle von automatisierter Kriegsführung. Am Schluss geht der Vortrag noch auf die Verantwortung von IT-Fachleuten ein, die ja das Wissen und Verständnis dieser Systeme mitbringen und daher überhaupt erst problematisieren können, wenn Systeme erweiterte oder gänzlich andere Funktionen erfüllen, als öffentlich und politisch oft kommuniziert und diskutiert wird. Überlegungen zu Handlungsoptionen und Auswegen leiten zuletzt die Diskussion ein. Licensed to the public under http://creativecommons.org/licenses/by/4.0 about this event: https://events.ccc.de/congress/2025/hub/event/detail/programmierte-kriegsverbrechen-uber-ki-systeme-im-kriegseinsatz-in-gaza-und-warum-it-fachleute-sich-dazu-auern-mussen

Dec 29, 20251h 0m

Making the Magic Leap past NVIDIA's secure bootchain and breaking some Tesla Autopilots along the way (39c3)

The Tegra X2 is an SoC used in devices such as the Magic Leap One, and Tesla's Autopilot 2 & 2.5 promising a secure bootchain. But how secure really is the secure boot? In this talk I go over how I went from a secured Magic Leap One headset, to exploiting the bootloader over USB, to doing fault injection to dump the BootROM, to finding and exploiting an unpatchable vulnerability in the BootROM's USB recovery mode affecting all Tegra X2s. In mid 2024, a friend approached me about Magic Leap making their TX2 based XR headsets little more than a paperweight by disabling the mandatory activation servers. I morally dislike this, companies shouldn't turn functional devices into e-waste just because they want to sell newer devices. After obtaining one, and poking at the Fastboot implementation, I discovered it was based off NVIDIA's Fastboot implementation, which is source available. I found a vulnerability in the NVIDIA provided source code in how it unpacks SparseFS images (named sparsehax), and successfully blindly exploited the modified implementation on the Magic Leap One. I also found a vulnerability in it that allowed gaining persistence via how it loads the kernel DTB (named dtbhax). Still unsatisfied with this, I used fault injection to dump the BootROM from a Tegra X2 devkit. In the BootROM I discovered a vulnerability in the USB recovery mode. Exploiting this vulnerability proved difficult due to only having access to memory from the perspective of the USB controller. I will explain what was tried, why it didn't work, and how I eventually got code execution at the highest privilege level via it. As I will demonstrate, this exploit also functions on Tesla's autopilot hardware. Licensed to the public under http://creativecommons.org/licenses/by/4.0 about this event: https://events.ccc.de/congress/2025/hub/event/detail/making-the-magic-leap-past-nvidia-s-secure-bootchain-and-breaking-some-tesla-autopilots-along-the-way

Dec 29, 202557 min

Supplements und Social Media – wenn der Online-Hype zur realen Gesundheitsgefahr wird (39c3)

Nicht zuletzt durch die Werbung in den sozialen Medien werden in Deutschland immer mehr Nahrungsergänzungsmittel verkauft. Einige Influencer bringen sogar ihre eigenen Präparate auf den Markt. Gleichzeitig häufen sich Fälle, in denen die Einnahme von vermeintlich harmlosen „Supplements“ zu Gesundheitsschäden geführt hat. Der Vortrag will daher die Mechanismen hinter dem Supplement-Hype aufzeigen, zudem erklären, warum aktuell ein ausreichender Verbraucherschutz insbesondere im Internet nicht gewährleistet werden kann, wo Handlungsbedarf für die Politik besteht und wie man sich selbst vor fragwürdigen Produkten schützen kann. Der Markt für Nahrungsergänzungsmittel boomt seit Jahren. Dafür sorgen unter anderem verschiedenste Influencer, die die Präparate in den sozialen Medien bewerben. Statt nur Produkte der großen Player in diesem Bereich anzupreisen, wie More Nutrition, ESN oder Holy Energy, haben einige Influencer mittlerweile sogar ihre eigenen Nahrungsergänzungsmittelmarken auf den Markt gebracht. Versprochen wird dabei vieles: Pre-Workout-Booster sollen die Leistung beim Krafttraining erhöhen und blitzschnell zum Traumkörper verhelfen, während Gaming-Booster Wachheit und eine Top-Performance beim Zocken versprechen. Wieder andere Kapseln oder auch Gummibärchen sollen für eine makellose Haut oder einen ruhigen Schlaf sorgen. Manche Präparate können angeblich sogar Krankheiten vorbeugen oder heilen. Doch was steckt tatsächlich in diesen Mitteln, die online regelrecht gehypt werden? Rein rechtlich handelt es sich um Lebensmittel, was wiederum bedeutet, dass sie ohne behördliche Zulassung auf den Markt gebracht werden dürfen. Es genügt schon, wenn der Unternehmer für die Sicherheit garantiert. Die Hürden für einen Marktzutritt sind damit denkbar niedrig, während gleichzeitig Gewinnmargen locken, die sogar den illegalen Drogenhandel übertreffen. Das Ergebnis zeigt sich in den Berichten der amtlichen Lebensmittelüberwachung: Bei den Proben, die das Niedersächsische Landesamt für Verbraucherschutz und Lebensmittelsicherheit im Jahr 2024 untersucht hat, entsprachen rund neun von zehn Proben (89 %) nicht den rechtlichen Vorgaben. Neben Mängeln bei der Kennzeichnung und Bewerbung, wodurch Verbraucher viel Geld für wirkungslose Pulver ausgeben, ist die stoffliche Zusammensetzung der Produkte besonders kritisch. So kann beispielsweise die Einnahme von überdosierten Vitamin-D-Präparaten zu Störungen des Calciumstoffwechsels führen (sog. Hypercalcämien). Vermeintlich harmlose pflanzliche Präparate, wie Kurkuma oder Ashwaganda, können zu Leberschäden bis hin zum Leberversagen führen. Besonders brisant ist dabei, dass die Wahrscheinlichkeit für die Erforderlichkeit einer Lebertransplantation oder den Tod des Patienten höher ist als bei Leberschäden durch Arzneimittel (83 vs. 66 %). Es kommen also Menschen durch die Einnahme von Präparaten zu Schaden, mit deren Hilfe sie ihrer Gesundheit eigentlich etwas Gutes tun wollten. Der Vortrag beleuchtet daher die aktuelle Marktsituation unter besonderer Berücksichtigung des Influencer-Marketings kritisch, erklärt den Unterschied zwischen Nahrungsergänzungs- und Arzneimitteln und stellt die rechtlichen Rahmenbedingungen für das Inverkehrbringen und die Bewerbung von Nahrungsergänzungsmitteln dar. Zudem wird aufgezeigt, warum ein ausreichender Verbraucherschutz durch die aktuellen Möglichkeiten des Lebensmittelrechts insbesondere im Internet nicht gewährleistet werden kann, wo Handlungsbedarf für die Politik besteht und wie man sich selbst vor fragwürdigen Produkten schützen kann. Licensed to the public under http://creativecommons.org/licenses/by/4.0 about this event: https://events.ccc.de/congress/2025/hub/event/detail/supplements-und-social-media-wenn-der-online-hype-zur-realen-gesundheitsgefahr-wird

Dec 29, 202558 min

Watch Your Kids: Inside a Children's Smartwatch (39c3)

Join us as we hack at a popular children's smartwatch and expose the secrets of every fifth child in Norway, their parents, and millions more. Smartwatches for children have entered the mainstream: Advertised on the subway and sold by your cell provider, manufacturers are charging premium prices comparable to an entry-level Apple watch. In exchange, parents are promised peace of mind: A safe, gentle introduction into the world of technology — and a way to call, text, and locate their child at any time. But how much are the vendor's promises of safety, privacy, GDPR compliance, apps made in Europe and cloud servers in Germany actually worth? We take you along the process of hacking one of the most popular children's watches out there, from gaining initial access to running our own code on the watch. Along the way, we find critical security issues at every turn. Our PoC attacks allow us to read and write messages, virtually abduct arbitrary children, and take control over any given watch. Finally, we'll also talk about disclosure, funny ideas of what passes as a security fix, and how we can use what we found to build something better. Licensed to the public under http://creativecommons.org/licenses/by/4.0 about this event: https://events.ccc.de/congress/2025/hub/event/detail/watch-your-kids-inside-a-children-s-smartwatch

Dec 29, 202536 min

Wer hat Angst vor dem Neutralitätsgebot? (39c3)

Wer überhaupt „neutral“ sein muss, was das bedeutet, und wer sich jetzt wehren muss. „Neutralität“ wird zum neuen Kampfbegriff: Weil sie gegen die menschenfeindliche Politik von Friedrich Merz protestieren, wirft die CDU Gruppen wie Omas gegen Rechts, Greenpeace und Correctiv vor, nicht neutral zu sein. Unter Berufung auf ein angeblich verletztes Neutralitätsgebot werden staatliche Förderungen gestrichen und NGOs geraten unter Beobachtung des Verfassungsschutzes. Julia Klöckner verbietet im Namen der „Neutralität“ Palestine-Shirts, Anstecknadeln und Regenbogenflaggen im Parlament. Die AfD fordert dazu auf, Lehrkräfte zu melden, die sich gegen Rechtsextremismus einsetzen oder entsprechende Positionen innerhalb der AfD kritisieren. Doch was steckt dahinter? Was bedeutet das sogenannte Neutralitätsgebot – und für wen gilt es überhaupt? Und für wen gilt es nicht? Zivilcourage kann nicht neutral sein – und soll es auch nicht sein. Genauso wie AfD-Hetze gegen Migrant*innen nicht „neutral“ ist, ist die Kritik menschenfeindlicher Äußerungen nicht nur legitim, sondern Pflicht demokratischer Bürger*innen. Das Beschwören eines „Neutralitätsgebots“ für NGOs ist ein durchschaubarer, aber gefährlicher Versuch, sie der eigenen Position zu unterwerfen. Die Rechtsanwältinnen Vivian Kube und Hannah Vos erklären den verfassungsrechtlichen Hintergrund, zeigen die autoritären Strategien hinter dem Ruf nach „Neutralität“ auf und geben Tipps, wie man sich dagegen wehren kann. Sie engagieren sich im Projekt Gegenrechtschutz, um demokratische Prinzipien und Betroffene vor rechtlichen Angriffen zu verteidigen. Licensed to the public under http://creativecommons.org/licenses/by/4.0 about this event: https://events.ccc.de/congress/2025/hub/event/detail/wer-hat-angst-vor-dem-neutralitatsgebot

Dec 29, 202542 min

When 8 Bits is Overkill: Making Blinkenlights with a 1-bit CPU (39c3)

Dec 29, 202531 min

Wer hat Angst vor dem Neutralitätsgebot? (39c3)

Dec 29, 202542 min

Shit for Future: turning human shit into a climate solution (39c3)

Dec 29, 202540 min

Watch Your Kids: Inside a Children's Smartwatch (39c3)

Dec 29, 202536 min

Shit for Future: turning human shit into a climate solution (39c3)

Humanity has already crossed the point where simply reducing emissions will no longer be enough to keep global warming below 2°C. According to the IPCC (AR6, WGIII), it is now essential to actively remove greenhouse gases from the atmosphere in order to meet global climate targets, maintain net-zero (or even net-negative emissions), and address the burden of historical emissions. At the same time, degraded soils and the climate crisis are a threat to global food security. Two years ago, I presented an overview of different methods available for carbon dioxide removal. Today, I want to show you an example of how CO₂ can be removed from the atmosphere while simultaneously improving the lives of local communities: Human shit. Human shit is a high abundant biomass, contains critical nutrients for global food security, and causes serious health and environmental issues from poor or non-existent treatment outside industrial countries. Converting shit into biochar presents a powerful solution: the process eliminates contaminants, stabilizes and locks away carbon, and can be used to improve agricultural soils. The challenge is that most nutrients in this biochar are not accessible to plants. To overcome this, I mixed human and chicken shit and produced a “Superchar” that releases far more nutrients. It’s not magic, it’s just some chemistry and putting aside your prejudices and disgust. I’ll show you how I did some shit experiments in Hamburg and Guatemala and how you can do it too. Today’s science mostly follows worn-out pathways and lack big discoveries and innovations. Scientists often don’t want to take a risk because the competition for a permanent position in academia is so high, which pressures them into conservative research topics supported by their supervisors. Even when science provides helpful solutions for urgent problems, the knowledge mostly ends up in libraries, written in papers that nobody understands. I want to show that it is worthwhile to follow research ideas that are unconventional, upset your boss af and explore topics that are unpopular like working with shit. I hope that sharing stories of how a funny idea turned into a solution encourage others to start making impact in their environment. Licensed to the public under http://creativecommons.org/licenses/by/4.0 about this event: https://events.ccc.de/congress/2025/hub/event/detail/shit-for-future-turning-human-shit-into-a-climate-solution

Dec 29, 202540 min

When 8 Bits is Overkill: Making Blinkenlights with a 1-bit CPU (39c3)

Over the last half year I have explored the Motorola mc14500 - a CPU with a true one-bit architecture - and made it simulate Conway's Game of Life. This talk gives a look into how implementing a design for such a simplistic CPU can work, and how it's possible to address 256 LEDs and half a kiloword of memory with just four bits of address space. In the late seventies, Motorola created a very cheap CPU, intended to replace logic circuits made from electromechanical relays. The resulting IC is so minimalistic that it can hardly be recognized as a CPU: Its data bus is just a single bit wide, it has no program counter, and the address bus isn't connected to the cpu at all. Yet, with just a few support components, and some clever programming, it can be made to do all sorts of things. We'll explore hardware design and programming by taking a look at my implementation of Conway's Game of Life, and answer the question of how one can address 512 words of memory, as well as some other peripherals, using just four bits of address space. Outline: * History and theory of operation of the mc14500 * Writing programs that process one bit at a time * A closer look at the hardware I built, including its wacky peripherals * Demonstration * Q&A Licensed to the public under http://creativecommons.org/licenses/by/4.0 about this event: https://events.ccc.de/congress/2025/hub/event/detail/when-8-bits-is-overkill-making-blinkenlights-with-a-1-bit-cpu

Dec 29, 202531 min

Zellkultur (39c3)

Ich bin ja glaube ich reichlich spät, aber da Moepern endlich wieder zum Congress kommt und ich das erst vor kurzem erfahren habe, dachte ich, ich werfe unseren Namen mal in den Hut. Wir sind/waren ein Biologie-Podcast und durch Babypause bei Moepern ist der etwas eingeschlafen. Schaffen wir ein Revival? Verratet ihr's uns. Wir nehmen auch den Podcast-Tisch wenns sonst nicht passt. :) In der Zellkultur reden Claudia (moepern) und Anna (Adora Belle) über das Leben, die Biologie und den ganzen Rest. Am liebsten schauen wir tief in die Zelle, auf die DNA und schauen Viren bei der Arbeit zu. Wir besprechen aktuelle Themen - aber wollen auch Grundlagenwissen vermitteln. Ihr seid außerdem eingeladen, eure Fragen zu stellen. Ob live oder per Zusendung. :) Und in einer aktuellen Bio-Frage klären wir Biologie-Mythen auf! Licensed to the public under http://creativecommons.org/licenses/by/4.0 about this event: https://events.ccc.de/congress/2025/hub/event/detail/zellkultur

Dec 29, 202545 min

Zellkultur (39c3)

Dec 29, 202545 min

BE Modded: Exploring and hacking the Vital Bracelet ecosystem (39c3)

The Vital Bracelet series is an ecosystem of interactive fitness toys, content on memory chips, and apps that talk via NFC. In this talk, we'll explore the hardware and software of the series, from its obscure CPU architecture, to how it interacts with the outside world, from dumping OTP ROMs and breaking security, to making custom firmware. The Vital Bracelet series, active from 2021 to 2024, was a line of toys that revolved around a number of fitness bracelets that encouraged exercise by raising characters from the Digimon series, and expanding into tokusatsu and popular anime characters later. Think of it as Tamagotchi, but nurturing through exercise instead of button presses. In this presentation, we'll look at the different parts of this series' ecosystem, how they work, and the different ways to circumvent various security measures and customize the devices' behavior. We start by looking at the first Vital Bracelet, with a quick introduction to hardware reverse engineering and how to dump firmware out of flash. Following that, we will take a look at the microcontroller used in the devices, and its obscure instruction set architecture. This will lead into an exploration of how to reverse engineer code when you are missing a significant portion of it, and how the embedded ROM was dumped. After this, we will look at the DRM applied to content, and how it was circumvented. Next, the device's NFC capabilities will be explored. With the release of the Vital Bracelet BE, which introduced upgradable firmware, came new challenges and opportunities. We will take a look at the new content format and additional DRM measures it incorporated, plus how the device's bootloader was dumped despite its signature verification scheme. Finally, we will take a look at the process for modding the various Vital Bracelet releases, and some techniques to use while writing patches. The material in this talk can be applied beyond just the Vital Bracelet series, and can be useful if you want to explore other electronic toys, or just hardware reverse engineering in general. Licensed to the public under http://creativecommons.org/licenses/by/4.0 about this event: https://events.ccc.de/congress/2025/hub/event/detail/be-modded-exploring-and-hacking-the-vital-bracelet-ecosystem

Dec 29, 202534 min

Teckids – eine verstehbare (digitale) Welt (39c3)

Die Teckids-Gemeinschaft bringt Kinder, Jugendliche und Erwachsene zusammen, um gemeinsam aktiv für eine verstehbare (digitale) Welt zu sein. Bei Teckids geht es nicht "nur" um Technikbasteln und Programmieren mit Kindern, sondern darum, mit anderen, für andere, bei Events und gesellschaftlich aktiv zu werden. In letzter Zeit haben wir viele Projekte dafür unternommen. Unter anderem haben wir den neuen Themen-Slot "Jung und überwacht" bei den BigBrotherAwards 2025 gestaltet und bereiten Jugendthemen für das nächste Jahr vor. Zum zweiten Mal laden wir beim 39c3 Kinder beim Fairydust-Türöffner-Tag "hinter die Kulissen" der Chaos-Teams ein. Unser Slogan mit dem etwas merkwürdigen Wort "Verstehbarkeit" steht dafür, dass alle nicht nur die Fähigkeiten, sondern auch das Recht behalten sollen, mit ihrer Technik zu machen, was sie wollen, und alles zu hinterfragen und zu verstehen. Dafür wollen wir noch mehr junge Menschen und auch Erwachsene erreichen. Licensed to the public under http://creativecommons.org/licenses/by/4.0 about this event: https://events.ccc.de/congress/2025/hub/event/detail/teckids-eine-verstehbare-digitale-welt

Dec 29, 202529 min

Teckids – eine verstehbare (digitale) Welt (39c3)

Dec 29, 202529 min

What Makes Bike-Sharing Work? (39c3)

Dec 29, 202540 min

BE Modded: Exploring and hacking the Vital Bracelet ecosystem (39c3)

Dec 29, 202534 min

What Makes Bike-Sharing Work? (39c3)

Bike- and e-bike-sharing promise sustainable, equitable mobility - but what makes these systems successful? Despite hundreds of cities operating thousands of shared bikes, trip data is rarely public. To address this, we built a geospatial analysis pipeline that reconstructs trip data from publicly accessible system status feeds. Using this method, we gathered **43 million km** of bike-sharing trips across **268 European cities**. Combined with over **100 urban indicators** per city, our analyses reveal how infrastructure, climate, demographics, operations, and politics shape system performance. We uncover surprising insights - such as why some e-bike systems underperform despite strong demand - and highlight how cities can design smarter, fairer mobility. All data and code are open-source, with an interactive demo at [dataviz.nefton.de](https://dataviz.nefton.de/). We are Felix, Georg, and Martin - each of us working professionally in different research and data areas, ranging from the future of mobility to computational fluid dynamics and machine learning. What unites us is our shared interest in **quantitative traffic analyses**. Building on earlier small-scale studies focused on individual cities, we set out to launch a project that captures shared bike system data across Europe - from regular bikes to e-bikes. In our study, which led to an **[open-access scientific publication](https://doi.org/10.1007/s11116-025-10661-2)**, we scraped shared bike data across Europe at a **minute-by-minute level** over many months, accumulating **more than 43 million records**. We analyze **behavioural and systemic patterns** to understand what makes a bike-sharing system useful and successful within a city. As such, this evidence-based research fits very well with the **39C3 Science track** and the theme of "**Power Cycles**" as we dissect the complex energy and usage cycles that define urban mobility and sustainable futures for everyone. We bridge the gap between urban planning, socioeconomics, and technology by applying statistical modeling and engineering knowledge to a large-scale mined dataset. Join us to learn whether right-wing politics stall sustainable mobility, or which climate e-bikes feel most comfortable in! We love going the extra mile and therefore provide a live, interactive demo that everyone can use to explore and understand traffic flows: [dataviz.nefton.de](https://dataviz.nefton.de). Therefore, attendees will be able to play with the data in a self-service way. We also provide all code on GitHub and the complete dataset on HuggingFace. And, of course, we will also discuss how both bike-sharing operators and our boss reacted when we told them about the dataset we already had collected (spoiler: lawyers were involved, yet it’s still available for downloads…). Licensed to the public under http://creativecommons.org/licenses/by/4.0 about this event: https://events.ccc.de/congress/2025/hub/event/detail/what-makes-bike-sharing-work-insights-from-43-million-kilometers-of-european-cycling-data

Dec 29, 202540 min

Hacking Karlsruhe - 10 years later (39c3)

"Wir gehen nach Karlsruhe!“ – das klang vor zehn Jahren nach Aufbruch und juristischem Hack. Heute ist klar: Strategische Prozessführung ist kein Sprint, sondern ein zähes, manchmal frustrierendes Dauerprojekt. In diesem Talk ziehen wir Bilanz: Was haben wir mit zivilgesellschaftlichen Verfassungsbeschwerden im Bereich Technologie erreicht – und wo sind wir gescheitert? Welche Fehler würden wir heute vermeiden, welche Wege waren richtig? Und was bedeutet es, wenn das höchste deutsche Gericht zunehmend weniger Lust auf digitalpolitische Grundrechtsfragen zeigt? Ein realistischer Blick hinter die Kulissen strategischer Klagen – und die Frage: Wie hackt man das Rechtssystem im Jahr 2025? Wenn Gesetze Grundrechte verletzen, warum nicht das Bundesverfassungsgericht hacken – mit Strategie, Teamwork und guter Begründung? Aus dieser Idee ist inzwischen ein zentrales Werkzeug zivilgesellschaftlicher Gegenmacht geworden: Strategische Prozessführung. Das Prinzip ist einfach: Gesetze nicht nur kritisieren, sondern systematisch angreifen, mit gezielten Verfassungsbeschwerden gegen Überwachung, Zensur und staatliche Eingriffe in die digitale Freiheit. Seitdem hat sich viel getan. Organisationen wie die Gesellschaft für Freiheitsrechte (GFF) haben den Weg nach Karlsruhe professionalisiert und Verfahren angestoßen, die viele aus den Nachrichten kennen: gegen die Vorratsdatenspeicherung, gegen das BND-Gesetz zur Auslandsüberwachung, gegen den Einsatz von Palantir, und gegen den Einsatz von Staatstrojanern. Einige dieser Verfahren waren erfolgreich und haben Gesetze gekippt. Andere sind krachend gescheitert – oder hängen seit Jahren in Karlsruhe fest. Dabei zeigt sich: Der Weg zum Urteil wird härter, die Erfolgsaussichten kleiner, und das Verfassungsgericht ist nicht mehr der progressive Motor, der es mal war. Dieser Talk zieht eine ehrliche Bilanz: Was bringt strategische Prozessführung wirklich? Was lässt sich aus Erfolgen und Misserfolgen lernen? Welche Fälle lohnen sich – und wo wird der Rechtsweg zur Sackgasse? Und wie verschiebt sich das Ganze inzwischen auf die europäische Ebene – wo neue Schauplätze wie der Digital Services Act oder der AI Act warten? Keine juristische Vorlesung, sondern ein Erfahrungsbericht aus zehn Jahren digitaler Grundrechtsarbeit. Es geht um Taktik, Fehlentscheidungen, unerwartete Allianzen – und um die Frage, wie man auch heute noch im Rechtssystem rütteln kann, wenn die Türen in Karlsruhe enger werden. Der Vortrag wird gehalten von Simone Ruf und Jürgen Bering von der Gesellschaft für Freiheitsrechte. Licensed to the public under http://creativecommons.org/licenses/by/4.0 about this event: https://events.ccc.de/congress/2025/hub/event/detail/hacking-karlsruhe-10-years-later

Dec 29, 202540 min

a media-almost-archaeology on data that is too dirty for "AI" (39c3)

Dec 29, 202537 min

Celestial navigation with very little math (39c3)

Dec 29, 202536 min

Hacking Karlsruhe - 10 years later (39c3)

Dec 29, 202540 min

a media-almost-archaeology on data that is too dirty for "AI" (39c3)

when datasets are scaled up to the volume of (partial) internet, together with the idea that scale will average out the noise, large dataset builders came up with a human-not-in-the-loop, cheaper-than-cheap-labor method to clean the datasets: heuristic filtering. Heuristics in this context are basically a set of rules came up by the engineers with their imagination and estimation to work best for their perspective of “cleaning”. Most datasets use heuristics adopted from existing ones, then add some extra filtering rules for specific characteristics of the datasets. I would like to invite you to have a taste together of these silent, anonymous yet upheld estimations and not-guaranteed rationalities in current sociotechnical artifacts, and on for whom these estimations are good-enough, as it will soon be part our technological infrastructures. In 1980s, non-white women’s body size data was categorized as dirty data when establishing the first women's sizing system in US. Now in the age of GPT, what is considered as dirty data and how are they removed from massive training materials? Datasets nowadays for training large models have been expanded to the volume of (partial) internet, with the idea of “scale averages out noise”, these datasets were scaled up by scrabbling whatever available data on the internet for free then “cleaned” with a human-not-in-the-loop, cheaper-than-cheap-labor method: heuristic filtering. Heuristics in this context are basically a set of rules came up by the engineers with their imagination and estimation that are “good enough” to remove “dirty data” of their perspective, not guaranteed to be optimal, perfect, or rational. The talk will show some intriguing patterns of “dirty data” from 23 extraction-based datasets, like how NSFW gradually equals to NSFTM (not safe for training model), and reflect on these silent, anonymous yet upheld estimations and not-guaranteed rationalities in current sociotechnical artifacts, and ask for whom these estimations are good-enough, as it will soon be part our technological infrastructures. Licensed to the public under http://creativecommons.org/licenses/by/4.0 about this event: https://events.ccc.de/congress/2025/hub/event/detail/a-media-almost-archaeology-on-data-that-is-too-dirty-for-ai

Dec 29, 202537 min

Celestial navigation with very little math (39c3)

Learn how to find your position using a sextant and a custom slide rule, almost no math required! Since the dawn of time people have asked themselves: where am I and why am I here? This talk won't help you answer the why question, but it will discuss how determine the where in the pre-GPS age of sextants, slide rules and stopwatches by taking the noon sight, aka the meridian passage. The usual way to find your position using the Sun requires a large almanac of lookup tables and some challenging math. The books are frustrating to consult on every sight and the base 60 degree-minute-second math is frustrating even with a calculator, and if you're on a traditional ship it seems wrong to do traditional navigation with electronic devices. To speed up the process I’ve designed a specialized circular slide rule that handles most of the table lookups to correct height of eye, semi-diameter, temperature, refraction and index errors, and also simplifies the degree-minute-second arithmetic required to calculate the exact declination of the Sun. In this talk I’ll demonstrate how to make your own printable paper slide rule and use it to reduce the meridian passage measurement to a lat/lon with just a few rotations of the wheels and pointer, no electronics or bulky books necessary! Licensed to the public under http://creativecommons.org/licenses/by/4.0 about this event: https://events.ccc.de/congress/2025/hub/event/detail/celestial-navigation-with-very-little-math

Dec 29, 202536 min

The Museum of Care: Open-Source Survival Kit Collection (39c3)

The talk is about the ideas behind setting up the David Graeber Institute and the Museum of Care. The Survival Kit Collection brings together collectives developing open source "social technologies" —spirulina farms, self-replicating 3D printers, modular housing, low-cost water systems, and ... art and education. In 2019, together with David Graeber, we held the first workshop about the Museum of Care at CCC to reimagine the relation between freedom, technology and value. Over these 6 years, the Museum of Care and the David Graeber Institute have experimented with various projects: the survival collection, Visual Assembly, and creating an open space for horizontal knowledge production—something we hope to develop into an actual University. We think humanity could already be living in a society of abundance and communal luxury. We have the technologies to produce enough for everyone to have everything. The issue isn't technological but social. This is why we need a Museum (of Care): museums are among the few places that create, distribute, and preserve what a society values. What will be at the session: We'll tell in more detail about the concept of the Museum of Care on abandoned ships (of which, according to Maritime Foundation data, there are more than 4,500 in the world). We'll talk about the halls of our museum: the Hall of Giants and other emerging spaces. Projects we're building—spirulina farms, 3D printers—in Saint Vincent (Caribbean) and Kibera Art District, Nairobi Kenya, Playground designed that communities can construct with nearly no resources. Can we actually build a nomadic museum proud not of its unique exhibits but of how easily they spread and get replicated? Then we will move to an open conversation about what poetic technologies are and how they differ from bureaucratic ones. Some people may have read David Graeber's book The Utopia of Rules; here you can download his other texts that are less widely known or not yet published. We would very much like to explore the question of poetic and bureaucratic technologies together with you. To facilitate this discussion, the David Graeber Institute has invited Alistair Parvin, creator of the Wiki House project, to join Nika Dubrovsky in conversation. The discussion continues in the format of a Visual Assembly—focused on building a distributed, non-hierarchical, genuinely open University with different ideas of funding and knowledge production. This is the very beginning of the process so all input is very much welcome. We'd welcome any ideas, critiques, or proposals for collaboration. Licensed to the public under http://creativecommons.org/licenses/by/4.0 about this event: https://events.ccc.de/congress/2025/hub/event/detail/the-museum-of-care-open-source-survival-kit-collection

Dec 29, 202537 min

Greenhouse Gas Emission Data (39c3)

Data about greenhouse gas emissions, both from countries and individual factories, is often publicly available. However, the data sources are often not as accessible and reliable as they should be. EU emission databases contain obvious flaws, and nobody wants to be responsible. Which factory in my city is the largest emitter of CO2? Which industrial sector is responsible for the largest share of a country's contribution to climate change? It should not be difficult to answer these questions. Public databases and reporting required by international agreements usually allow us to access this data. However, trying to access and work with these datasets — or, shall we say, Excel tables — can be frustrating. UN web pages that prevent easy downloads with a "security firewall", barely usable frontends, and other issues make it needlessly difficult to gain transparency about the sources of climate pollution. While working with official EU datasets, the speaker observed data points that could not possibly be true. Factories suddenly dropped their emissions by orders of magnitude without any explanation, different official sources report diverging numbers for the same emission source, and responsible European and National authorities appear not to care that much. The talk will show how to work with relevant greenhouse gas emission data sources and how we can access them more easily by converting them to standard SQL tables. Furthermore, we will dig into some of the strange issues one may find while investigating emission datasets. # Background / Links * Why is it needlessly difficult to access UNFCCC Emission Data? https://industrydecarbonization.com/news/why-is-it-needlessly-difficult-to-access-unfccc-emission-data.html * UNFCCC Emission Data Downloads: https://industrydecarbonization.com/docs/unfccc/ * Code (Docker, MariaDB/MySQL, phpMyAdmin) to easily access EU emisison data: https://github.com/decarbonizenews/ghgsql Licensed to the public under http://creativecommons.org/licenses/by/4.0 about this event: https://events.ccc.de/congress/2025/hub/event/detail/greenhouse-gas-emission-data-public-difficult-to-access-and-not-always-correct

Dec 29, 202539 min

The Museum of Care: Open-Source Survival Kit Collection (39c3)

Dec 29, 202537 min

Design for 3D-Printing (39c3)

Dec 29, 202537 min

Lightning Talks - Tag 3 (39c3)

Dec 29, 20251h 59m

Lightning Talks - Tag 3 (39c3)

Lightning Talks - Tag 3 - **Lightning Talks Introduction** - **"Oma, erzähl mir von der Zukunft" oder: Wie wir weiter interessante Sachen machen, ohne den Planeten zu ruinieren 🌱** — *EstherD* - **Don't abuse the ecosystem** — *michele* - **The Climatepoetry.org video tool** — *Magnus Ahltorp* - **Neo-Kolonialismus & Katzenbilder - Installation zur Lieferkette von GenAI** — *Stefan, Yannik & Rike* - **Build social inventories with StashSphere** — *Maximilian Güntner* - **Invitation to the Fermentation Camp "Kvas 2026"** — *algoldor* - **Stretching nginx to its limits: a music player in the config file** — *Eloy* - **2D Graphics Creation with Graphite - How to build a hackable graphics Editor** — *Dennis Kobert* - **The Modulator: a Custom Controller for Live Music Performance** — *Jakob Kilian* - **Find hot electronic devices for cheap using Lock-In Thermography** — *Clemens Grünewald* - **Those Who Control** — *Andreas Haupt* - **SearchWing - Search&Rescue Drones** — *searchwing team* - **Hacking ID3 MP3 Metadata** — *Danilo Erazo* - **Genetic engineering with CRISPR/Cas9: how far are we today from biopunk?** — *Dmytro Danylchuk* - **Discovering the Orphan Source Village** — *Martin Hamilton* - **kicoil - generate planar coils in any shape for PCBs and ICs** — *jaseg* - **Trade Offer: Pentest Data for CTF Points** — *Sebastian* - **Soziologische Gabentheorie - Grundlage für die Bewertung von Social Media?** — *sozialwelten* - **WissKomm Wiki - Bibliothek für Videos und Podcasts** — *TimBorgNetzWerk* - **Shitty Robots** — *Neo* - **Lightning** — *Vi* Licensed to the public under http://creativecommons.org/licenses/by/4.0 about this event: https://events.ccc.de/congress/2025/hub/event/detail/lightning-talks-tag-3

Dec 29, 20251h 59m

Design for 3D-Printing (39c3)

3D-Printers have given us all the unprecedented ability to manufacture mechanical parts with a very low barrier to entry. The only thing between your idea and its physical manifestation is the process of designing the parts. However, this is actually a topic of incredible depth: Design engineering is a whole discipline to itself, built on top of tons and tons of heuristics to produce shapes that are functional, strong, and importantly: well-manufacturable In this talk, I will present the rules for designing well-printable parts and touch on other areas of design considerations so you can learn to create parts that work first try and can be reproduced by others on their 3d-printers easily. Over the years, the 3d-printing community has discovered many tricks and rules that help creating parts that can be printed well and fulfill their purpose as best as possible. I started collecting these rules and wrote an article guide to make this knowledge more accessible. I want to present the most important principles and the mindset that is needed to achieve perfected design. This is not about how to use a CAD program to design a part — but rather about the thought process of the design engineer while drawing up a part. A though process that consists of compromises between many objectives, of heuristic rules, and many neat little tricks. The article that this talk is based on can be found on my blog: https://blog.rahix.de/design-for-3d-printing/ Licensed to the public under http://creativecommons.org/licenses/by/4.0 about this event: https://events.ccc.de/congress/2025/hub/event/detail/design-for-3d-printing

Dec 29, 202537 min

FAFO: How we stopped worrying and bought an Electron Microscope (39c3)

We went on Kleinanzeigen and started a non-profit semiconductor lab. We created a space to do the kind of research we've always wanted, but couldn't do at our local hackerspaces - either due to space constraints, or lack of structures allowing for a safe operation of a proper lab. -- This video was independently recorded & produced by dmi about this event: https://events.ccc.de/congress/2025/hub/en/event/detail/fafo-how-we-stopped-worrying-and-bought-an-electro

Dec 29, 20251h 1m

GPTDash – Der Reverse-Turing-Test (39c3)

KIs (bzw. LLMs) wirken immer menschlicher. Schon längst ist es schwer bis unmöglich zu erkennen, ob ein Text von einer KI oder einem Menschen geschrieben wurde. Maschinen dringen immer mehr in den menschlichen Diskurs ein. Wir wollen das nicht länger hinnehmen und drehen den Spieß um. In unserem Reverse-Turing-Test schlüpfen die Teilnehmenden in die Rolle einer KI und versuchen so robotisch-menschlich wie möglich zu klingen. In einer anschließenden Blindstudie prüfen wir, wer sich am besten unter KIs mischen und beim nächsten Robot Uprising die Spionin der Wahl wäre. Humor, Kreativität und ein Hang zu allgemeingültigen, nichtssagenden Floskeln sind die perfekten Voraussetzungen! Ein digitales Endgerät (Smartphone, Tablet, Laptop, …) reicht zum Mitspielen aus. Licensed to the public under http://creativecommons.org/licenses/by/4.0 about this event: https://events.ccc.de/congress/2025/hub/event/detail/gptdash-der-reverse-turing-test

Dec 29, 20251h 31m

Code to Craft: Procedural Generation for the Physical World (39c3)

Join bleeptrack for a deep dive into the fascinating world of procedural generation beyond the screen. From stickers and paper lanterns to PCBs, furniture, and even physical procedural generators, this talk explores the challenges and creative possibilities of bringing generative projects into tangible form. In this talk, I will share practical insights from developing procedural generation tools for physical objects: ranging from stickers and paper lanterns to printed circuit boards and even furniture. I will outline key challenges and considerations when generating designs for fabrication tools such as laser cutters or pen plotters, as well as how to adapt procedural systems so they can be reproduced by a wide audience (not everyone has access to CNC machines or industrial equipment, sadly!). Beyond technical considerations, I aim to encourage attendees to translate their own generative ideas into tangible artifacts and to foster a culture of open-sourcing and knowledge sharing within the community. Licensed to the public under http://creativecommons.org/licenses/by/4.0 about this event: https://events.ccc.de/congress/2025/hub/event/detail/code-to-craft-procedural-generation-for-the-physical-world

Dec 28, 202535 min

The Small Packet of Bits That Can Save (or Destabilize) a City (39c3)

Dec 28, 202540 min

Reverse engineering the Pixel TitanM2 firmware (39c3)

Dec 28, 202537 min

Code to Craft: Procedural Generation for the Physical World (39c3)

Dec 28, 202535 min

Reverse engineering the Pixel TitanM2 firmware (39c3)

The TitanM2 chip has been central to the security of the google pixel series since the Pixel 6. It is based on a modified RISC-V design with a bignum accelerator. Google added some non standard instructions to the RISC-V ISA. This talk investigates the reverse engineering using Ghidra, and simulation of the firmware in python. I will discuss the problems encountered while reverse engineering and simulating the firmware for the TitanM2 security chip, found in the Google Pixel phones. I'll discuss how to obtain the firmware. Talk about the problems reverse engineering this particular binary. I show how you can easily extend ghidra with new instructions to get a full decompilation. Also, I wrote a Risc-V simulator in python for running the titanM2 firmware. Licensed to the public under http://creativecommons.org/licenses/by/4.0 about this event: https://events.ccc.de/congress/2025/hub/event/detail/reverse-engineering-the-pixel-titanm2-firmware

Dec 28, 202537 min

The Small Packet of Bits That Can Save (or Destabilize) a City (39c3)

The Emergency Alert System (EAS) and its SAME (Specific Area Message Encoding) protocol are public alerting technologies that broadcast short digital bursts over VHF triggering emergency messages on millions of receivers across North America. In Mexico, this technology was integrated into the Seismic Alert System (SASMEX) which more than 30 million people in the central part of the country rely on to prepare for frequent earthquakes. While new alerting technologies have emerged, the EAS-SAME network continues to play an important role for public safety in the U.S., Mexico, and Canada. Yet, the same small packets of bits that can help protect a city could also, in the wrong hands, destabilize it. This talk examines how these systems operate and reveals a troubling truth: spoofing these alerts is far easier than most people expect. In this talk, we’ll begin by contextualizing the importance of the seismic alert in Mexico City, a system born from the devastating 1985 earthquake. We’ll examine how it was designed, how it works, and why it carries such a deep psychological impact. From there, we’ll explore the history and design of Weather Radio and the SAME protocol, looking at how messages are transmitted and encoded through this technology, and how it was later adapted for SASMEX. I’ll also share my personal experience building compatible receivers, from early open-source experiments that inspired local manufacturers to create government-certified devices, to developing a receiver as part of my undergraduate thesis. We’ll analyze how simplicity, one of the key strengths of these systems, also introduces certain risks, and how these trade-offs emerge when dealing with accessibility, interoperability, and security in system design. Finally, I’ll demonstrate how to receive, decode, and encode these alert messages, and discuss how, with the right equipment, it’s possible to generate such alert signals. Licensed to the public under http://creativecommons.org/licenses/by/4.0 about this event: https://events.ccc.de/congress/2025/hub/event/detail/the-small-packet-of-bits-that-can-save-or-destabilize-a-city

Dec 28, 202540 min

51 Ways to Spell the Image Giraffe: (39c3)

Generative AI models don't operate on human languages – they speak in **tokens**. Tokens are computational fragments that deconstruct language into subword units, stored in large dictionaries. These tokens encode not only language but also political ideologies, corporate interests, and cultural biases even before model training begins. Social media handles like *realdonaldtrump*, brand names like *louisvuitton*, or even *!!!!!!!!!!!!!!!!* exist as single tokens, while other words remain fragmented. Through various artistic and adversarial experiments, we demonstrate that tokenization is a political act that determines what can be represented and how images become computable through language. Tokens are the fragments of words that generative models use to process language, the step that breaks text into subword units before any neural networks are involved. There are 51 ways to combine tokens to spell the word giraffe using existing vocabulary: from a single token **giraffe** to splits using multiple tokens like *gi|ra|ffe*, *gira|f|fe*, or even *g|i|r|af|fe*. In one experiment, we hijacked the prompting process and fed token combinations directly to text-to-image models. With variations like *g|iraffe* or *gir|affe* still generating recognizable results, our experiments show that the beginning and end of tokens hold particular semantic weight in forming giraffe-like images. This reveals that certain images cannot be generated through prompting alone, as the tokenization process sanitizes most combinations, suggesting that English, or any human language, is merely a subset of token languages. The talk features experiments using genetic algorithms to reverse-engineer prompts from images, respelling words in token language to change their generative outcomes, and critically examining token dictionaries to investigate edge cases where the vocabulary breaks down entirely, producing somewhat *speculative languages* that include strange words formed at the edge of chaos where English meets token (non-)sense. These experiments show that even before generation occurs, token dictionaries already encode a stochastic worldview, shaped by the statistical frequencies of their training data – dominated by popular culture, brands, platform-speak, and *non-words*. Tokenization is, therefore, a political act: it defines what can be represented and how the world becomes computationally representable. We will look at specific tokens and ask: Which models use which vocabularies? What *non-word* tokens are shared among models? And how do language models make sense of a world using a language we do not understand? Licensed to the public under http://creativecommons.org/licenses/by/4.0 about this event: https://events.ccc.de/congress/2025/hub/event/detail/51-ways-to-spell-the-image-giraffe-the-hidden-politics-of-token-languages-in-generative-ai

Dec 28, 202538 min

Xous: A Pure-Rust Rethink of the Embedded Operating System (39c3)

Xous is a message-passing microkernel implemented in pure Rust, targeting secure embedded applications. This talk covers three novel aspects of the OS: hardware MMU support (and why we had to make our own chip to get this feature), how and why we implemented the Rust standard library in Rust (instead of calling the C standard library, like most other Rust platforms), and how we combine the power of Rust semantics with virtual memory to create safe yet efficient asynchronous messaging primitives. We conclude with a short demo of the OS running on a new chip, the "Baochip-1x", which is an affordable, mostly-open RTL SoC built in 22nm TSMC, configured expressly for running Xous. The world is full of small, Internet-of-Things (IoT) gadgets running embedded operating systems. These devices generally fall into two categories: larger devices running a full operating system using an MMU which generally means Linux, or smaller devices running without an MMU using operating systems like Zephyr, chibios, or rt-thread, or run with no operating system at all. The software that underpins these projects is written in C with coarse hardware memory protection at best. As a result, these embedded OSes lack the security guarantees and/or ergonomics offered by modern languages and best practices. The Xous microkernel borrows concepts from heavier operating systems to modernize the embedded space. The open source OS is written in pure Rust with minimal dependencies and an emphasis on modularity and simplicity, such that a technically-savvy individual can audit the code base in a reasonable period of time. This talk covers three novel aspects of the OS: its incorporation of hardware memory virtualization, its pure-Rust standard library, and its message passing architecture. Desktop OSes such as Linux require a hardware MMU to virtualize memory. We explain how ARM has tricked us into accepting that MMUs are hardware-intensive features only to be found on more expensive “application” CPUs, thus creating a vicious cycle where cheaper devices are forced to be less safe. Thanks to the open nature of RISC-V, we are able to break ARM’s yoke and incorporate well-established MMU-based memory protection into embedded hardware, giving us security-first features such as process isolation and encrypted swap memory. In order to make Xous on real hardware more accessible, we introduce the Baochip-1x, an affordable, mostly-open RTL 22nm SoC configured expressly for the purpose of running Xous. The Baochip-1x features a Vexriscv CPU running at 400MHz, 2MiB of SRAM, 4MiB of nonvolatile RRAM, and a quad-core RV32E-derivative I/O accelerator called the “BIO”, based on the PicoRV clocked at 800MHz. Most Rust targets delegate crucial tasks such as memory allocation, networking, and threading to the underlying operating system’s C standard library. We want strong memory safety guarantees all the way down to the memory allocator and task scheduler, so for Xous we implemented our standard library in pure Rust. Adhering to pure Rust also makes cross-compilation and cross-platform development a breeze, since there are no special compiler or linker concerns. We will show you how to raise the standard for “Pure Rust” by implementing a custom libstd. Xous combines the power of page-based virtual memory and Rust’s strong borrow-checker semantics to create a safe and efficient method for asynchronous message passing between processes. This inter-process communication model allows for easy separation of different tasks while keeping the core kernel small. This process maps well onto the Rust "Borrow / Mutable Borrow / Move" concept and treats object passing as an IPC primitive. We will demonstrate how this works natively and give examples of how to map common programming algorithms to shuttle data safely between processes, as well as give examples of how we implement features such as scheduling and synchronization primitive entirely in user space. We conclude with a short demo of Xous running on the Baochip-1x, bringing Xous from the realm of emulation and FPGAs into everyday-user accessible physical silicon. Licensed to the public under http://creativecommons.org/licenses/by/4.0 about this event: https://events.ccc.de/congress/2025/hub/event/detail/xous-a-pure-rust-rethink-of-the-embedded-operating-system

Dec 28, 202539 min