AllYouCanTech

“Most Dogs Take Part in Fetch, While Cats Join in More Often Than Anticipated”

### The Unexpected Realm of Fetch: A Comparison of Cats and Dogs

When we envision the traditional game of fetch, we often picture eager dogs dashing across a yard, zealously retrieving a ball or stick. In contrast, cats frequently get a reputation for being detached and self-sufficient, seeming more inclined to lazily bat at a toy rather than partake in a fetch game. Nonetheless, recent studies are overturning these ideas, indicating that a notable proportion of cats also find joy in fetch, suggesting that this behavior is more widespread in both species than was formerly assumed.

#### Cats and Their Fetching Behaviors: A Fresh Insight

An intriguing study recently released in *PLoS ONE* has provided fresh insights into the fetching activities of cats, demonstrating that nearly 41% of pet cats engage in this playful endeavor. This statistic is surprising, considering the widely held assumption that cats show little interest in games led by humans. The research, co-authored by Mikel Delgado, a researcher in veterinary medicine at Purdue University, underscores that fetching is not solely a canine activity but is also relished by a significant number of felines.

Delgado and her colleagues were fascinated by the scarcity of research on fetching behavior in dogs, and even more so by the limited studies focusing on cats. Their results imply that cats, often seen as solitary and independent creatures, can be quite social and interactive with their human friends. This challenges the stereotype of the detached cat, revealing that numerous felines are not just willing but enthusiastic about participating in fetch games.

#### Playful Behavior Across Different Species

Playful behavior is prevalent in various animal species, particularly amongst mammals and birds. While dogs are well-recognized for their playful disposition, cats also showcase an array of playful actions, many of which imitate hunting behavior. These include chasing, pouncing, and stalking—actions commonly observed in their wild counterparts like European wildcats and lynxes.

Interestingly, kittens typically engage in more social play, wrestling with their siblings, before they shift to more solitary play in adulthood. This contrasts with dogs, who often commence with solitary play and later partake in more sociable activities. Despite these distinctions, both species demonstrate a capability for interactive play with humans, including the game of fetch.

#### What Drives a Cat to Fetch?

A 2023 study by British researchers explored the intricacies of fetching behavior in cats through an online questionnaire targeted at cat owners. The findings indicated that most cats who enjoy fetch developed this behavior independently, without specific training. Additionally, cats often take control of the game, playing longer and retrieving more frequently when they initiate the activity.

The research also highlighted that some cat breeds are more predisposed to partake in fetching. Breeds like Siamese, Burmese, Tonkinese, and Bengal cats were among those most likely to join in. These breeds are distinguished by their high energy levels and playful characteristics, potentially explaining their inclination toward this behavior.

#### Dogs and Fetch: A Familiar Connection

While the revelation of fetching behavior in cats is captivating, dogs remain the clear leaders in fetching. The very study that examined cat behavior also investigated dogs, discovering that 77.8% of dogs engage in fetch. Specific breeds, such as Labradors, Golden Retrievers, Border Collies, and English Cocker Spaniels, are particularly drawn to fetching, likely due to their trainability and the roles they have historically fulfilled in human society.

Interestingly, the study revealed that male dogs are more inclined to play fetch than females, and older dogs are more prone to engaging in the activity. However, dogs cohabiting with other dogs are less likely to fetch, possibly due to the presence of alternative social interactions.

#### What Accounts for the Difference?

The authors of the study propose that the variance in fetching behavior between cats and dogs may stem from their respective domestication histories. Dogs have been selectively bred for specific roles, such as hunting and herding, which often require retrieving objects. This might have fostered a stronger instinct for fetching in dogs. Conversely, cats have not undergone selective breeding for such purposes, making their fetching behavior more unexpected.

Delgado suggests that the more pertinent inquiry is not why dogs fetch more frequently, but rather why so many cats do, given that they haven’t been bred for this activity. The answer could lie in the playful disposition of cats and their ability for social engagement, qualities that are often undervalued.

### Conclusion

The activity of fetch, traditionally viewed as a dog’s realm, is more prevalent within the animal kingdom than previously recognized. With nearly 41% of cats participating in this activity as well, it’s apparent that our feline companions are more playful and interactive than their image suggests. As research continues to delve into the behaviors of our pets, we may discover that cats and dogs share more similarities in their playful capacities than we ever imagined. Whether you prefer dogs, cats, or both, it’s evident that the connection between humans and their pets is

Installing a Ratgdo on My Garage Door: Reasons I Regret Not Doing It Earlier

### The Emergence of DIY Home Automation: An Experience with the Ratgdo

In an era where smart homes dominate, allowing control over everything from lights to refrigerators with a simple voice command, it’s no wonder that garage doors have embraced the automation trend as well. Yet, not all options are equal, and for individuals who prioritize privacy, functionality, and user-friendliness, finding the ideal garage door automation tool can prove challenging. Enter the **Ratgdo**, a compact yet powerful device that has captured the admiration of DIY aficionados and privacy-minded homeowners.

#### The Drawbacks of OEM Solutions

For numerous homeowners, the initial step towards smart garage solutions is often the original equipment manufacturer (OEM) products. Take, for instance, the LiftMaster garage door opener, part of the Chamberlain Group, which promotes the **MyQ** system as its primary automation option. However, many users have realized that MyQ is quite flawed.

The MyQ system requires an extra accessory to work with Apple’s HomeKit, yet this accessory is no longer available or supported by Chamberlain. Additionally, the MyQ app, essential for operating the garage door, has raised alarms regarding data privacy. Its privacy policy implies that user data might be exploited for profit, a significant concern for those wishing to keep their personal information secure.

Due to these limitations, numerous homeowners, including the writer of this narrative, have sought alternatives that provide enhanced functionality while safeguarding privacy.

#### Introducing the Ratgdo

The **Ratgdo** represents a third-party solution that overcomes the restrictions of OEM systems like MyQ. Created by Paul Weiland, the Ratgdo is a compact, Wi-Fi-enabled device that connects directly to the terminals of your garage door opener. It operates independently of any cloud services, ensuring that your data remains within your home network, and it integrates effortlessly with popular home automation platforms like Apple’s HomeKit, Google Home, and Home Assistant.

The Ratgdo’s name, short for “Rage Against the Garage Door Opener,” reflects its defiant spirit. It is crafted for those disillusioned with the constraints and privacy issues of OEM products who are eager to seize control of their home automation.

#### Configuring the Ratgdo

One of the most attractive features of the Ratgdo is its straightforward setup. The process starts with installing the correct firmware on the device. For HomeKit users, this means connecting the Ratgdo to a computer via USB and adhering to the instructions found on a dedicated webpage. After the firmware is uploaded, the device can link to your home Wi-Fi network and be incorporated into your HomeKit setup with a simple QR code scan.

The actual installation is just as simple. The Ratgdo connects to your garage door opener’s terminals and is placed between the door sensors and the opener itself. Power is supplied through a micro-USB connection, with the device including a USB adapter for convenience. The entire installation process, including wiring and mounting, takes only a few minutes.

#### The Satisfaction of Functional Home Automation

Upon installation, the Ratgdo fulfills its promise of effortless home automation. With just a voice command, you can open or shut your garage door, monitor its status, and even weave it into larger home automation routines. For instance, you could create a routine that automatically closes the garage door when you leave or opens it upon your return.

The Ratgdo’s local operation ensures that your data stays within your home network, providing reassurance for privacy-conscious users. Furthermore, its compatibility with various home automation platforms makes it a flexible solution that can evolve alongside your home automation needs.

#### The Future of Home Automation

The Ratgdo exemplifies the potential of DIY home automation. It’s an affordable, compact device that delivers a level of functionality and privacy often absent in OEM solutions. For those willing to invest some effort, the Ratgdo offers a smart, secure, and fulfilling way to automate their garage door.

However, like any DIY approach, there are considerations. The Ratgdo’s compatibility is confined to certain garage door opener models, and it requires different firmware based on the home automation platform being utilized. Nonetheless, for those comfortable with a little experimentation, these are minor obstacles in the quest for a truly intelligent home.

In a landscape where privacy is increasingly at risk, the Ratgdo provides a revitalizing alternative to cloud-dependent solutions. It serves as a reminder that often the best solutions are those you create yourself. In the case of the Ratgdo, it is a solution that meets your exact needs without compromising.

So, if you’re fed up with peering out the window to ensure your garage door is shut, or if you’re hesitant to surrender your personal data to an anonymous corporation, the

DOJ Subpoenas Nvidia as AI Antitrust Probe Intensifies, Report Indicates

### Nvidia’s Antitrust Investigation: An In-Depth Look at the DOJ’s Inquiry and Its Consequences for the AI Sector

The Department of Justice (DOJ) has markedly escalated its scrutiny of Nvidia, the tech powerhouse renowned for its supremacy in the AI computing domain. What started as an initial assessment has transformed into a comprehensive antitrust investigation, with the DOJ issuing subpoenas to Nvidia and various other tech entities. The inquiry aims to determine if Nvidia is exploiting its market power to partake in anti-competitive behaviors, possibly inhibiting competition within the swiftly advancing AI sector.

#### Origin of the Investigation

The DOJ’s examination of Nvidia first surfaced in June 2024, when it came to light that the agency was probing Nvidia’s extraordinary dominance—estimated to be around 90%—in the market for chips that can power AI models. These chips are vital for training and executing intricate AI algorithms, making them essential for firms eager to harness artificial intelligence’s capabilities.

At first, industry analysts speculated that the investigation might not have a substantial impact on Nvidia’s operations. Market reactions reflected this, as Nvidia’s stock remained fairly stable after the preliminary reports. However, the dynamics shifted dramatically in September 2024 when Bloomberg disclosed that the DOJ had progressed beyond mere inquiries into competitors and started issuing subpoenas. This revelation led to a significant sell-off, resulting in Nvidia experiencing a staggering $279 billion decline in market value in a single day—the largest recorded single-day drop in market capitalization.

#### Claims of Anti-Competitive Conduct

Sources affiliated with the inquiry have indicated that the DOJ is especially worried that Nvidia could be partaking in practices that disadvantage both competitors and consumers. A central accusation is that Nvidia may be providing preferential supply and pricing to customers that exclusively utilize its technology or buy its comprehensive systems. This practice could potentially bind customers to Nvidia’s ecosystem, making it challenging for them to transition to alternative suppliers.

Another aspect of the investigation revolves around Nvidia’s acquisition of RunAI, a company focused on AI infrastructure. The DOJ suspects this acquisition could further cement Nvidia’s dominance by compelling RunAI clients to use Nvidia chips, thereby constraining their choices and hindering competition.

#### The Wider Context: AI Sector Under Examination

Nvidia is not the sole company under the DOJ’s watchful eye. The agency is also investigating various components of the AI sector, such as computing power, data utilized for training large language models, cloud service providers, and access to critical hardware like GPUs. Jonathan Kanter, the DOJ’s antitrust chief, has stressed the significance of intervening “in real time” to avert any single entity from monopolizing the AI landscape.

This broader examination follows the DOJ’s recent success in establishing that Google has a monopoly in the search engine arena. The agency seems determined to thwart a comparable situation from emerging in the AI field, where Nvidia’s advanced AI chips are perceived as a “scarce resource.”

#### Nvidia’s Reaction and Market Perspective

In light of the DOJ’s intensifying inquiry, Nvidia has insisted that its achievements are rooted in merit. The company highlights its benchmark outcomes and the value it delivers to clients as proof that it is not engaged in anti-competitive activities. Nvidia’s CEO, Jensen Huang, has also indicated that the company prioritizes clients who can immediately utilize its products, especially in ready-to-deploy data centers, to avoid stockpiling its highly sought-after AI chips.

Despite the recent downturn in its stock value, Nvidia’s market leadership appears stable for now. The company’s shares have more than doubled in 2024, and its ecosystem comprises nearly 5 million developers and 40,000 businesses. Nvidia’s AI chips serve as the foundation for over 1,600 generative AI companies, and the organization is projected to finish the year with profits surpassing the total revenue of its nearest rival, AMD.

#### Possible Challenges to Nvidia’s Supremacy

While Nvidia currently holds a commanding position in the AI market, some analysts posit that its hold could falter over time. A recent opinion piece in the Financial Times argued that Nvidia’s dominance may wane as the focus in AI shifts from training large “foundation” models to deploying these models in consumer and business applications. This transition could create opportunities for competitors that offer less powerful but more affordable chips, particularly in the inference chip sector.

Nvidia is already encountering considerable competition in this area, not only from startups but also from major tech players such as Meta and Amazon. Furthermore, global regulatory agencies are closely monitoring Nvidia’s operations. For example, French competition regulators raided Nvidia’s offices last year as part of a larger inquiry into competition in cloud computing.

#### Future Directions: DOJ’s Upcoming Actions

The DOJ’s investigation into Nvidia is anticipated to persist for several months, if not longer. Former assistant attorney general Bill Baer has noted that the DOJ’s objective is to foresee how Nvidia’s existing market dominance could negatively impact the industry in the

Australian Government Research Shows AI Falls Short When Compared to Humans in Summarization Tasks

### The Difficulties of AI-Produced Summaries: Takeaways from ASIC’s Proof-of-Concept Research

As large language models (LLMs) gain traction, their possible applications are being examined in diverse sectors. One of the most notable functions of LLMs is their capability to swiftly condense lengthy documents, enhancing information accessibility and comprehension for users. Nonetheless, a recent analysis by Australia’s Securities and Investments Commission (ASIC) has pointed out significant constraints in the existing functionalities of these models, especially when placed against human-generated summaries.

#### ASIC’s Research: An In-Depth Investigation

In a proof-of-concept examination carried out in partnership with Amazon Web Services (AWS), ASIC scrutinized the summarization abilities of various LLMs, including the Llama2-70B model. The objective was to evaluate how effectively these models could summarize public submissions to an external Parliamentary Joint Committee inquiry focusing on audit and consultancy firms. A strong summary, according to ASIC, should underline references to ASIC, recommendations for mitigating conflicts of interest, and requests for enhanced regulation, while also providing page citations and brief contextual clarifications.

The study’s results were not particularly hopeful for advocates of AI-generated summaries. The Llama2-70B model, despite being among the more extensive and sophisticated models available at that time, produced summaries that were considered significantly inferior to those created by human specialists. The AI-generated summaries were critiqued for being “redundant and irrelevant,” frequently merely echoing the original submissions without providing significant context or analysis.

#### Main Observations: Where AI Lacks

The ASIC research unveiled multiple pivotal shortcomings in AI-generated summaries:

1. **Deficiency of Nuance and Contextual Grasp**: A prominent challenge was the AI’s poor capacity to dissect and summarize intricate content requiring a profound understanding of context, subtlety, or implied meanings. This shortcoming resulted in summaries that felt generic and failed to reflect the specific mentions of ASIC within the submissions.

2. **Errors and Irrelevance**: The AI-generated summaries were also noted to contain inaccuracies, miss pertinent details, and emphasize inconsequential points. This problem was intensified by occurrences of AI hallucinations—cases where the model created grammatically correct but factually wrong text.

3. **Increased Demand on Human Resources**: Instead of streamlining the summarization effort, the AI-generated outputs were likely to burden human reviewers with more work. The necessity to verify the AI’s summaries and the observation that the original material often conveyed information more effectively indicated that the AI’s participation could be counterproductive.

#### The Impact of Model Size and Prompt Design

While focusing primarily on the Llama2-70B model, ASIC also examined smaller models like Mistral-7B and MistralLite during initial stages. The comparison affirmed the notion within the industry that larger models generally yield superior outcomes. However, ASIC’s results also highlighted the critical role of prompt design—carefully constructing the inquiries and tasks posed to the model—to attain the best results. Moreover, behind-the-curtain adjustments to model parameters, such as temperature and top-k sampling, were required to refine the outputs.

#### Constraints and Future Outlook

It’s crucial to understand that ASIC’s study had various limitations that hinder broad application of the findings across all LLMs or scenarios. For instance, the researchers had merely one week to optimize their model, and they believe that dedicating more time to this phase could lead to enhanced outcomes. Additionally, the research centered on the Llama2-70B model, which has now been outclassed by more advanced models like ChatGPT-4o, Claude 3.5 Sonnet, and Llama3.1-405B. These newer models might provide better performance, particularly in tasks necessitating a deep comprehension of context and nuance.

Despite the unsatisfactory findings, ASIC holds a hopeful perspective on the future of generative AI. The organization recognizes that advancements in this domain are rapid, and upcoming models are expected to enhance both efficacy and precision. Nevertheless, the study acts as a cautionary reminder for large enterprises contemplating the integration of LLMs into their operations. It underscores the significance of comprehensive evaluation and optimization prior to depending on AI-generated outputs, especially in tasks demanding high levels of accuracy and context understanding.

#### Conclusion

ASIC’s proof-of-concept study offers essential insights into the present limitations of AI-generated summaries. Although LLMs such as Llama2-70B display potential, they still struggle in areas where nuanced understanding and accurate contextual analysis are required. As the technology progresses, it is anticipated that future models will rectify some of these deficiencies. Yet, at this juncture, organizations should approach the implementation of AI-generated summaries with caution, ensuring that human oversight is an integral aspect of the procedure.

Zyxel Raises Security Warning for Various Product Vulnerabilities

### Zyxel Alerts Users to Serious Vulnerabilities in Networking Equipment: Urgent Action Needed

Zyxel, a manufacturer of networking hardware, has released a crucial security advisory, notifying users of nearly twelve vulnerabilities present in a broad array of its products. If not addressed, these weaknesses may enable attackers to gain full control over the affected devices, potentially using them as gateways into larger networks. In light of the criticality of these vulnerabilities, Zyxel is advising all users to implement the provided patches without delay.

#### The Most Significant Vulnerability: CVE-2024-7261

The most pressing of these vulnerabilities is designated as **CVE-2024-7261**, which has received an alarming severity score of 9.8 out of 10. This vulnerability allows an unauthenticated attacker to execute OS commands by transmitting a specially crafted cookie to a susceptible device. It arises from the improper handling of special elements in the ‘host’ parameter within the CGI program of the affected access points and security routers. Almost 30 models from Zyxel are at risk due to this vulnerability, making it crucial for users to promptly update their firmware.

#### Other Vulnerabilities in Firewall Models

Zyxel has also discovered seven additional vulnerabilities affecting its firewall models, such as the ATP, USG-FLEX, and USG FLEX 50(W)/USG20(W)-VPN. These vulnerabilities vary in severity, with ratings from 4.9 to 8.1, and include:

– **CVE-2024-6343**: A buffer overflow vulnerability in the CGI program that may enable an authenticated attacker with admin access to conduct a denial-of-service (DoS) attack through specially crafted HTTP requests.

– **CVE-2024-7203**: A post-authentication command injection vulnerability that could permit an authenticated attacker with administrator privileges to execute OS commands by running a specially designed CLI command.

– **CVE-2024-42057**: A command injection vulnerability in the IPSec VPN feature, which might allow an unauthenticated attacker to execute OS commands by submitting a crafted username. This attack is feasible only if the device employs User-Based-PSK authentication and a valid user with a username longer than 28 characters exists.

– **CVE-2024-42058**: A null pointer dereference vulnerability in select firewall versions that could enable an unauthenticated attacker to carry out DoS attacks by sending crafted packets.

– **CVE-2024-42059**: A post-authentication command injection vulnerability that permits an authenticated attacker with admin rights to execute OS commands by uploading a specially crafted compressed language file via FTP.

– **CVE-2024-42060**: Another post-authentication command injection vulnerability allowing an authenticated attacker with administrator access to run OS commands by uploading a crafted internal user agreement file to the vulnerable device.

– **CVE-2024-42061**: A reflected cross-site scripting (XSS) vulnerability in the CGI program “dynamic_script.cgi,” which could enable an attacker to deceive a user into accessing a crafted URL containing XSS payloads. If the malicious script runs, the attacker could gather browser-based information from the victim’s browser.

#### Buffer Overflow Vulnerability: CVE-2024-5412

Furthermore, **CVE-2024-5412** is another notable vulnerability discovered in 50 Zyxel product models, encompassing various customer premises equipment (CPE), fiber optical network terminals (ONTs), and security routers. This buffer overflow issue lies within the “libclinkc” library of the affected devices and could let an unauthenticated attacker launch DoS attacks via crafted HTTP requests. The severity score for this vulnerability is rated at 7.5.

#### Increasing Threat Landscape

Zyxel devices have increasingly become targets for cyber attackers in recent years. Numerous vulnerabilities in these devices have been actively utilized in the wild, some even facilitating large-scale Distributed Denial-of-Service (DDoS) assaults. For example, in 2021, hackers exploited a backdoor found in Zyxel devices, and in 2023, DDoS botnets continued to take advantage of Zyxel devices with known vulnerabilities.

#### Immediate Action Needed

Given the urgent nature of these vulnerabilities, Zyxel has made patches available for the majority of affected devices. Users can download these patches from the links provided within the security advisories. In certain instances, patches can be accessed through the cloud, while for a limited number of products, users may need to directly contact Zyxel’s support team to acquire the necessary updates.

**Users are highly encouraged to implement these patches immediately to safeguard their networks against potential attacks.**

#### Conclusion

The revelation of these vulnerabilities highlights the necessity for routine security updates and proactive network oversight. As networking devices remain a primary target for cybercriminals, being proactive about potential threats is essential. Zyxel’s prompt action in releasing patches is praiseworthy,