AllYouCanTech

Archaeologists Discover Possible Earliest Recognized Bronze Age Board Game

### The Ancient Game of *Hounds and Jackals*: A Recent Find in Azerbaijan

An age-old board game referred to as *Hounds and Jackals*, or *Fifty-Eight Holes*, has historically been linked to ancient Egypt, where it was thought to have originated. Nevertheless, a recent archaeological find in Azerbaijan indicates that the game may have roots in Asia, potentially predating its Egyptian versions. This insight, detailed in a paper published in the *European Journal of Archaeology*, calls into question long-standing beliefs about the game’s origins and underscores the significance of board games in ancient cultures.

#### A Game of Antiquity

*Hounds and Jackals* is a competitive race game akin to modern games such as backgammon. Players strive to advance their pieces—typically resembling hounds or jackals—across a board with two parallel rows of 29 holes. The goal is to be the first to navigate all pieces from the starting position to the larger hole at the board’s end. The game was initially identified by modern researchers in 1890, and since then, approximately 60 examples have emerged from sites in Egypt, Mesopotamia, Israel, Syria, Iran, and Azerbaijan.

The earliest known iteration of the game harks back to the era of Mentuhotep II (2064–1952 BCE) in Egypt, where it was a favored pastime among the elite. However, the recent finding in Azerbaijan implies that the game might have existed even earlier in Asia, prompting inquiries into its authentic origins.

#### The Discovery in Azerbaijan

In 2018, archaeologist Walter Crist, a co-author of the study, made a significant discovery at the Çapmalı rock shelter located near the western coast of the Caspian Sea in Azerbaijan. Crist and his team uncovered a game board etched into stone, exhibiting the recognizable pattern of 58 holes set in two parallel rows. This board bore a striking resemblance to other *Hounds and Jackals* boards unearthed in Egypt and southwestern Asia.

The Çapmalı site, along with nearby areas including Ağdaşdüzü, Yeni Türkan, and Dübəndi, produced pottery fragments and other artifacts indicating that the game was played by shepherds and locals during the Middle Bronze Age. The researchers propose that these communities engaged with the game as a leisure activity during the protracted winter months, finding refuge in rock formations like Çapmalı.

#### A Broader Context

The unearthing of the *Hounds and Jackals* board in Azerbaijan is part of a larger trend in archaeology that has revealed ancient board games from diverse parts of the world. For example, a 4,000-year-old board discovered in Oman’s Qumayrah Valley is thought to be a precursor to the *Royal Game of Ur*, another ancient race game that may have impacted the evolution of backgammon. In a similar vein, a 500-year-old board for the game *Mill* (also known as *Nine Men’s Morris*) was recently uncovered in the ruins of Ćmielów Castle in Poland, illustrating how ancient games surpassed cultural and geographical divides.

The distribution of *Hounds and Jackals* over such an extensive area—from Egypt to Mesopotamia and now Azerbaijan—attests to the game’s popularity and its function as a social instrument. The researchers posit that the game may have encouraged interactions among different social groups, acting as a “social lubricant” that fostered connections and communication, much like contemporary board games.

#### Challenges in Dating the Game

While the discovery in Azerbaijan is exhilarating, the researchers caution that accurately dating the game poses a challenge. The Middle Bronze Age context of the Çapmalı site implies that the game may have been played in the region before its introduction in Egypt, but additional evidence is necessary for confirmation. “Clearly, further early evidence for the game from precisely dated contexts is required to credit a specific culture for inventing the game,” the authors noted.

Regardless of its precise origins, the swift spread of *Hounds and Jackals* across such a broad area highlights the game’s cultural importance. It served not merely as a form of amusement but also as a means to nurture social cohesion and interaction among diverse societies.

#### Conclusion

The discovery of a *Hounds and Jackals* board in Azerbaijan enriches our comprehension of ancient board games and their significance in human history. While the game has long been connected to Egypt, this new finding suggests its beginnings might extend further east into Asia. As archaeologists persist in uncovering more data, we may soon gain further insights into how this ancient game traversed various cultures and what it reveals about the people who engaged with it.

Meanwhile, the game of *Hounds and Jackals* remains a captivating illustration of how ancient societies employed games not solely for entertainment but as a means of forging connections across time and space.

—

**European Journal of Archaeology**, 2024

Users Receive Ban Alerts for Trying to Investigate OpenAI’s Newest Model’s “Thoughts”

## OpenAI’s “Strawberry” AI Model: The Debate Surrounding Concealed Reasoning

In September 2024, OpenAI unveiled its newest series of AI models, under the code name “Strawberry,” which comprises the o1-preview and o1-mini models. These models were created to demonstrate superior reasoning capabilities, marking an advancement compared to earlier versions like GPT-4o. However, the launch has ignited considerable controversy, particularly regarding OpenAI’s choice to hide the underlying “chain of thought” that the models utilize to reach their conclusions. This has prompted a surge of intrigue, hacking efforts, and even alerts from OpenAI to users attempting to delve too deeply into the model’s mechanics.

### What is the “Chain of Thought”?

The “chain of thought” signifies the sequential reasoning methodology that the o1 models employ to address problems. In contrast to former models that delivered answers more directly, o1 has been specifically trained to decompose complex inquiries into smaller, logical components before concluding. This strategy aims to enhance the model’s efficacy in managing reasoning assignments, rendering it more dependable in tackling intricate issues.

In the ChatGPT interface, users can observe a filtered representation of this chain of thought under a “Thinking” subheader. Nevertheless, the unfiltered, raw reasoning process remains concealed from users. Instead, a secondary AI model interprets the raw chain of thought and offers a more refined, user-centric version. OpenAI has made it clear that this design choice was deliberate, but it has also incited a spike of interest from hackers and AI enthusiasts eager to reveal the concealed reasoning.

### The Quest to Unveil the Hidden Chain of Thought

Immediately after OpenAI’s launch of the o1 models, users commenced attempts to “jailbreak” or employ “prompt injection” techniques to deceive the AI into disclosing its raw chain of thought. These tactics involve altering the model’s inputs in ways that circumvent its safeguards, often by exploiting vulnerabilities in the prompt structure or the model’s internal logic. While there have been isolated incidents of success, none have been definitively verified.

OpenAI has been proactive in overseeing these endeavors through the ChatGPT interface. The company has reportedly issued alerts to users who attempt to investigate too intricately into the model’s reasoning mechanisms. Some users have even received emails from OpenAI, cautioning them that their actions breach the company’s terms of use and may lead to a ban from utilizing the o1 models.

### OpenAI’s Reaction: Warning Emails and Policy Enforcement

Numerous users have recounted their experiences of receiving warning emails from OpenAI after inquiring the o1 model about its reasoning procedures. For example, one user reported that merely using the phrase “reasoning trace” prompted a warning. Others have observed that even posing general inquiries regarding the model’s reasoning can elicit a similar reaction.

The warning emails generally indicate that the user’s request has been flagged for breaching OpenAI’s regulations against circumventing safeguards. The emails urge users to cease this activity and to engage with ChatGPT in alignment with the company’s terms of service. Repeated infractions could lead to loss of access to the o1 models, which OpenAI internally refers to as “GPT-4o with Reasoning.”

This crackdown has exasperated some within the AI community, especially those engaged in red-teaming and safety research. Marco Figueroa, who oversees Mozilla’s GenAI bug bounty programs, was among the first to publicly express his discontent after receiving a warning email. He contended that the limitations obstruct his capability to conduct beneficial safety research on the model, which could aid in identifying vulnerabilities and bolstering the system’s resilience.

### The Rationale Behind Concealing the Chain of Thought

In a blog post titled “

Google Backs Privately Financed Satellite Network for Prompt Wildfire Detection

### FireSat: Pioneering Wildfire Detection from Orbit

With the increasing occurrence and severity of wildfires driven by climate change, the necessity for swift detection and action has reached a critical level. Recent advancements in space technology are providing innovative solutions to tackle this escalating issue. While satellites have been utilized for monitoring wildfires, a groundbreaking project, **FireSat**, seeks to transform the methods of detecting and tracking these catastrophic events. Overseen by the nonprofit entity **Earth Fire Alliance (EFA)** and backed by significant philanthropic contributors, FireSat is preparing to launch a fleet of small satellites uniquely engineered for real-time wildfire detection.

### The Escalating Demand for Enhanced Wildfire Detection

While wildfires are a natural occurrence in many ecosystems, their prevalence and severity have surged in recent years, fueled by rising global temperatures and extended periods of drought. In 2021, for instance, the **Windy Fire** devastated regions of California’s Sequoia National Forest, endangering ancient sequoia groves. Such incidents underscore the pressing requirement for more timely and precise wildfire detection systems.

Historically, satellites have been deployed to track wildfires, yet these systems frequently lack the necessary sensitivity and speed to identify smaller blazes before they escalate. Enter FireSat, a cutting-edge project designed to recognize wildfires as small as a classroom anywhere on the planet.

### FireSat: A Revolutionary Approach to Fire Detection

FireSat will be a constellation comprising over 50 small satellites launched into **low-Earth orbit**. In contrast to conventional Earth-observing satellites, FireSat will be the inaugural fleet committed exclusively to the detection and monitoring of wildfires. The initiative has received partial funding from **Google’s philanthropic division**, which recently declared a new investment of **$13 million** to aid the endeavor. This financial support will facilitate adherence to the timeline for the inaugural launch of the FireSat pathfinder satellite, anticipated next year, with the complete constellation potentially operational by 2026.

The **Earth Fire Alliance** states that FireSat marks a considerable advancement in wildfire management. “As fires intensify and propagate more swiftly, we believe that collaborative efforts are essential for fostering necessary innovations in fire management and tackling climate change,” the organization articulated in a statement.

### State-of-the-Art Technology for Immediate Detection

The satellites for FireSat will be crafted by **Muon Space**, a California-based startup specializing in satellite development. Each satellite will be outfitted with **six-band multispectral infrared instruments**, capable of surveying an area of the Earth up to **900 miles (1,500 kilometers)** wide. These sophisticated sensors will possess the sensitivity to identify wildfires as small as **16 by 16 feet (5 by 5 meters)**, marking a significant leap from existing systems.

What distinguishes FireSat is its integration of **Google AI** for the swift analysis of satellite imagery to pinpoint potential wildfires. The AI will juxtapose new data with earlier images to catch flare-ups while also considering local weather conditions and surrounding infrastructure. This enables emergency responders to obtain immediate notifications regarding the location and magnitude of a fire, facilitating more rapid and effective actions.

“US Encounters Difficulties in Prohibiting TikTok Due to Security Issues While Ignoring Temu and Comparable Applications”

# TikTok vs. US Government: A Legal Confrontation Regarding National Security and Free Expression

The ongoing legal confrontation between TikTok and the US government has reached a pivotal moment. On September 16, 2024, the U.S. Court of Appeals conducted oral arguments in the case *TikTok Inc. v. Merrick Garland*, wherein TikTok, along with its parent company ByteDance and US-based creators, strives to avert a potential ban of the app in the United States. This case holds significant consequences, not only for TikTok’s prospects in the US but also for matters related to national security, free expression, and the extent of governmental oversight.

## The Central Controversy: National Security vs. Free Expression

At the core of the case lies legislation enacted by Congress that targets foreign adversaries, requiring TikTok to separate from its Chinese parent, ByteDance. The U.S. government asserts that TikTok’s ties to China present a considerable national security threat. The concern revolves around the possibility that ByteDance’s oversight of TikTok’s source code, which is updated in China, could enable the Chinese government to access sensitive information from American users and manipulate platform content to sway public discourse.

TikTok’s legal representatives, under the guidance of attorney Andrew J. Pincus, argue that the law is “unprecedented” and could yield “staggering” impacts on free expression, particularly affecting the 170 million Americans who utilize the app. Pincus argued before the three-judge panel that the statute contravenes the First Amendment as it specifically targets a US-based entity—TikTok US—without adequate evidence that the app is being exploited for malicious intent.

## The Government’s Position: A Threat to National Security

The U.S. government, represented by attorney Daniel Tenny, defended the legislation by underscoring the potential dangers posed by TikTok. Tenny claimed that the data collected by TikTok is “extremely valuable to a foreign adversary,” and that ByteDance’s control over the application’s source code opens the door for the Chinese government to manipulate content shown to American users. The government argues that this isn’t merely an issue of free expression, but rather one of national security.

Tenny further asserted that the content on TikTok is not entirely American, as it is influenced by “Chinese engineers in China” who command the app’s algorithms and source code. The government contends that the sole method to alleviate this risk is for ByteDance to divest from TikTok, thus removing any potential Chinese governmental influence.

## TikTok’s Counterargument: Insufficient Evidence and Overreach

TikTok’s defense is built upon the premise that there is no solid evidence bolstering the government’s assertions. Pincus highlighted that the FBI has merely issued alerts regarding the “potential” for the Chinese Communist Party to utilize TikTok for espionage or propaganda, yet there have been no confirmed instances of such behavior. He also criticized the law for being selectively applied, noting that other China-based applications like Temu and Shein, which also gather sensitive data, have not faced similar scrutiny.

Pincus cautioned that if the court validates the law, it could create a perilous precedent, permitting the government to regulate speech across any platform based on foreign ownership. This, he contended, would effectively invert the principles of the First Amendment.

## The Involvement of TikTok Creators

TikTok creators have also entered the legal fray, claiming that the law would impose unfair constraints on their capacity to express themselves and connect with audiences. Jeffrey Fisher, an attorney representing TikTok creators, underscored that TikTok differs from other social media platforms. He emphasized that numerous creators have cultivated substantial followings on TikTok that they cannot replicate elsewhere, such as YouTube or Instagram. Fisher asserted that mandating TikTok’s divestment or closure would fundamentally change the nature of the content and tools available to creators, thus infringing upon their rights to free expression.

Fisher further emphasized the distinctive bond between TikTok creators and their audiences, arguing that Americans possess a “fundamental interest” in selecting the platform for their expressions. He proposed that the law is driven by an intention to suppress expression, making it hard for the government to provide justification under the First Amendment.

## The Court’s Challenge: Balancing Security and Free Expression

Throughout the proceedings, the three-judge panel appeared to wrestle with the intricate issues at stake. On one side, the government maintains a legitimate interest in safeguarding national security, particularly in an age where data holds significant value. Conversely, the court must weigh whether the law disproportionately targets TikTok and its users, thereby infringing upon their constitutional rights.

One of the pivotal inquiries posed by the panel pertained to whether the law would remain permissible in wartime conditions. Pincus acknowledged that under wartime circumstances, the government might possess greater latitude to regulate speech,

Boar’s Head Halts Liverwurst Manufacturing After Fatal Outbreak Associated with 9 Deaths

### Boar’s Head Deli-Meat Facility Ceases Operations Due to Lethal Listeria Outbreak

A deli-meat facility of Boar’s Head in Jarratt, Virginia, has been permanently shut down in the wake of a lethal *Listeria monocytogenes* outbreak that has resulted in nine fatalities and affected 57 individuals across 18 states. The outbreak, traced back to the liverwurst production line at the plant, has prompted the recall of over 7 million pounds of Boar’s Head meat items. The company has since declared that it will stop liverwurst production and is implementing significant changes to enhance its food safety measures.

#### The Outbreak and Plant Shutdown

Maryland health investigators were the first to identify the outbreak, linking the *Listeria* strain to Boar’s Head liverwurst. This discovery initiated a nationwide recall of deli meats and led to the indefinite shutdown of the Jarratt facility in late July 2024. On September 13, Boar’s Head confirmed that the contamination stemmed from a specific production process exclusive to the liverwurst line at the Jarratt plant, resulting in the company’s decision to permanently halt liverwurst production.

While this announcement clarifies the outbreak’s origin, it also reveals more serious issues within the plant. Inspection reports uncovered numerous sanitation deficiencies, such as mold, water leaks, and unsanitary conditions that facilitated the spread of the deadly bacteria.

#### Sanitation Deficiencies

From August 2023 to August 2024, the Jarratt plant was cited for 69 violations, including serious sanitation failures that posed a considerable risk to public health. Inspection reports indicated the presence of mold, algal growth, and “meat buildup” on machinery. Inspectors also observed flies, gnats, and other insects, along with unpleasant odors and blood pooling on the floors.

A 2022 inspection had already flagged concerns, labeling the plant as an “imminent threat” to public health. During that inspection, numerous issues were identified, including extensive rust, deli meats exposed to wet ceilings, green mold, and wall holes. Despite these serious findings, the plant continued operations until the outbreak necessitated its closure.

#### Cross-Contamination and Ineffective Controls

A recent document from the USDA’s Food Safety and Inspection Service (FSIS) provides insights into the contamination’s spread within the facility. During an inspection on July 24 and 25, 2024, *Listeria* was detected on a pallet jack in a room designated for processing ready-to-eat meats. Although this room was deemed “low risk” for *Listeria* by Boar’s Head, the USDA concluded the company’s *Listeria* control measures were “ineffective.”

The contaminated pallet jack was employed to transport racks of Beechwood Hams between blast coolers and production lines. However, inspectors found that pallet jacks and racks were not confined to specific production lines, allowing cross-contamination among different meat types. Employees moved between production areas without changing their personal protective equipment (PPE), heightening the contamination risk.

Inspectors noted workers engaged in tasks like trash removal, floor cleaning, and maintenance without following proper hygiene protocols. These lapses contributed to *Listeria* spreading throughout the facility, impacting multiple products beyond just liverwurst.

#### The Human Impact

As of the most recent reports, 57 individuals have been hospitalized due to the outbreak, and nine have succumbed. The CDC has interviewed 44 affected persons, with 25 indicating they consumed deli-sliced liverwurst. However, others consumed different deli meats, implying that the contamination may have affected various products within the facility.

#### Boar’s Head’s Reaction and Future Initiatives

In response to the outbreak, Boar’s Head is dedicated to overhauling its safety and quality assurance programs across all its facilities. The company is recruiting seasoned food safety professionals and establishing new protocols to avert future outbreaks. In a statement, Boar’s Head affirmed its commitment to restoring consumer confidence, stating:

> “You have our promise that we will work tirelessly to regain your trust and ensure that all Boar’s Head products consistently meet the high standards that you deserve and expect. We are determined to learn from this experience and emerge stronger.”

While the company has initiated steps to manage the ongoing crisis, the long-term effect on its reputation remains uncertain. The closure of the Jarratt facility and the cessation of liverwurst production signify a notable transformation for Boar’s Head, a brand recognized for its high-quality deli meats.

#### What is Listeria?

*Listeria monocytogenes* is a bacterium that can

Widespread Problems with Secure Boot Flaw “PKfail” More Extensive Than Previously Acknowledged

# PKfail: A Supply Chain Compromise Threatening Secure Boot Across Thousands of Devices

In a major security incident impacting various computing equipment, a supply chain compromise has weakened Secure Boot defenses across multiple models, which include ATMs, point-of-sale (POS) devices, medical instruments, gaming consoles, and voting machines. This vulnerability, known as **PKfail**, results from the employment of non-production test platform keys that were never intended for live deployment. The issue has persisted for over ten years, influencing devices from prominent manufacturers such as Acer, Dell, HP, Lenovo, among others.

## What is Secure Boot?

Secure Boot is a security framework created to guarantee that a device boots strictly using software that is verified by the Original Equipment Manufacturer (OEM). It is a component of the Unified Extensible Firmware Interface (UEFI) and employs cryptographic keys to confirm the integrity of the firmware and bootloader prior to the operating system initiating. Should the software lack a pre-approved digital signature, Secure Boot will prevent it from executing, thereby stopping malicious code from running during the boot sequence.

The foundation of Secure Boot hinges on the **Platform Key (PK)**, a cryptographic key embedded within the system’s firmware. This key establishes the root-of-trust, forming a secure bond between the hardware and firmware. In principle, this guarantees that only verified software is operative on the device.

## The PKfail Dilemma

The PKfail vulnerability emerges from the application of **non-production test platform keys** in numerous device models. These keys, frequently labeled with notices like “DO NOT TRUST,” were never intended for production usage. Nonetheless, due to a collapse in the supply chain and security management, these test keys found their way into genuine products, undermining the entire Secure Boot mechanism.

The private segment of one of these test keys was even made public on GitHub in 2022, rendering it available to a broad audience. This data could facilitate advanced attacks, such as embedding **rootkits** into the UEFI firmware, granting attackers sustained control over a device while circumventing Secure Boot safeguards entirely.

## The Extent of the Issue

Initially, researchers from the security firm **Binarly** pinpointed approximately 513 device models utilizing compromised test keys. However, deeper analysis disclosed that the quantity of affected models is substantially larger—**972 models** at present. Additionally, the number of models utilizing the specific key released on GitHub has escalated from 215 to **490**. The researchers also identified **four additional test keys**, raising the total number of compromised keys to around 20.

The impacted devices cover a diverse array of sectors and application scenarios, including:

– **Medical instruments**
– **Gaming consoles**
– **ATMs**
– **Point-of-sale (POS) devices**
– **Enterprise servers**
– **Voting machines**

The vulnerability spans multiple vendors. Leading device manufacturers like **Acer, Dell, Gigabyte, HP, Lenovo, Intel, Supermicro, and Fujitsu** are involved. Moreover, smaller companies such as **Hardkernel, Beelink, and Minisforum** have also been found to utilize compromised keys in their offerings.

## The Supply Chain Challenge

The root of PKfail can be traced back to the intricacies of the contemporary hardware supply chain. Numerous device manufacturers depend on third-party suppliers for essential components like UEFI firmware. In this scenario, the test keys originated from **AMI**, one of the principal providers of UEFI software development kits (SDKs). Further insights have disclosed that other UEFI providers, including **Insyde** and **Phoenix**, are similarly implicated.

This extensive usage of non-production keys underscores a profound failure in supply chain security. As **Fabio Pagani**, a researcher at Binarly, noted, “The complexity of the supply chain is outweighing our capacity to effectively handle the risks tied to third-party suppliers.”

## The Consequences of PKfail

The PKfail vulnerability erodes the assurances offered by Secure Boot, a safeguard mandated for specific government contractors and required in numerous corporate settings. Secure Boot is also recognized as a best practice for individuals and organizations encountering high-risk threats, such as nation-state adversaries or advanced persistent threats (APTs).

For devices that rely upon Secure Boot, PKfail presents a considerable hazard. Attackers could exploit the compromised keys to bypass Secure Boot safeguards, permitting the installation of malware that remains even after system reboots or operating system reinstalls. This could generate severe repercussions, particularly in critical infrastructures like ATMs, medical devices, and voting machines.

However, for individuals or organizations that do not utilize Secure Boot, PKfail does not introduce an additional risk beyond the standard threats associated with firmware vulnerabilities.

## Detection and Remediation

In light of the PKfail revelation, Binarly has introduced a **free detection tool** that enables users to ascertain if