Connect with us

Artificial Intelligence

Track People and Read Through Walls with Wi-Fi Signals

blank

Published

on

blank

Recent research has shown that your Wi-Fi router’s signals can be used as a sneaky surveillance system to track people and read text through walls.

Recently, Carnegie Mellon University computer scientists developed a deep neural network that digitally maps human bodies using Wi-Fi signals.

It works like radar. Many sensors detect Wi-Fi radio waves reflected around the room by a person walking. This data is processed by a machine learning algorithm to create an accurate image of moving human bodies.

“The results of the study reveal that our model can estimate the dense pose of multiple subjects, with comparable performance to image-based approaches, by utilizing WiFi signals as the only input,” the researchers wrote in a December 2022 pre-print paper.

The team claims this experimental technology is “privacy-preserving” compared to a camera, despite concerns about intrusion. The algorithm can only detect rough body positions, not facial features and appearance, so it could provide a new way to monitor people anonymously.

They write, “This technology may be scaled to monitor the well-being of elder people or just identify suspicious behaviors at home.”

Recent research at the University of California Santa Barbara showed another way Wi-Fi signals can be used to spy through walls. They used similar technology to detect Wi-Fi signals through a building wall and reveal 3D alphabet letters.

WiFi still imagery is difficult due to motionlessness. “We then took a completely different approach to this challenging problem by tracing the edges of the objects,” said UC Santa Barbara electrical and computer engineering professor Yasamin Mostofi.

 

As Editor here at GeekReply, I'm a big fan of all things Geeky. Most of my contributions to the site are technology related, but I'm also a big fan of video games. My genres of choice include RPGs, MMOs, Grand Strategy, and Simulation. If I'm not chasing after the latest gear on my MMO of choice, I'm here at GeekReply reporting on the latest in Geek culture.

Continue Reading

Artificial Intelligence

Elon Musk intends to impose a fee on new X members in order to grant them the ability to make posts

blank

Published

on

blank

Elon Musk intends to impose a small fee on new X members to grant them the ability to make posts on the social network and to address the issue of automated accounts.

Musk responded to an X account that had mentioned modifications on X’s website by stating that imposing a little cost on new accounts was the sole method to prevent the overwhelming influx of automated programs, known as bots.

“The current state of artificial intelligence, including troll farms, can easily pass the ‘are you a bot’ test,” Musk stated, alluding to techniques such as CAPTCHA.

In response to another user, Musk then stated that newly created accounts would have the ability to post without incurring a fee after a period of three months.

Currently, there is a lack of information regarding the timeline for implementing this policy and the potential charges that may be imposed on new members, which is consistent with the pattern of limited disclosure often seen in announcements pertaining to the social network.

In October of last year, X implemented a policy of charging new members who have not been verified a fee of $1 per year in both New Zealand and the Philippines. Users that register for the site from these regions will have the ability to view the posts, but they will not be able to engage with them. In order to publish, endorse, share, respond to, save, and cite posts, a monetary charge was required. Musk may implement a charge that is comparable to fees in other areas.

In a recent announcement, X stated that the platform would be initiating a significant elimination of spam accounts. Users were cautioned that their follower count could be impacted. Nevertheless, by implementing a strategy to impose fees on new members, the social media corporation appears to be striving to address the issue of bots more effectively.

Although Musk has expressed concerns about AI bots, X recently updated its policy to state that public posts may be utilized for training machine learning algorithms or artificial intelligence models. In July 2023, Musk mentioned that his AI company, xAI, would utilize public posts for training models.

In recent weeks, xAI has introduced its Grok chatbot to premium users of X, who are charged a monthly fee of $8. The chatbot was previously accessible to users who subscribed to the Premium+ tier for $16 per month. Last week, Fortune reported that X intends to offer Grok to users for composing posts.

Continue Reading

Artificial Intelligence

According to insiders, Tesla has laid off high-performing employees and made significant cuts in certain departments

blank

Published

on

blank

According to a reliable source, Tesla management informed employees on Monday that the recent layoffs, which resulted in a significant reduction of staff in certain divisions, including high-performing individuals, were primarily a result of the company’s weak financial performance.

The team was informed of the layoffs only one week prior to Tesla’s anticipated release of its first-quarter earnings. Tesla’s decision to take this step is a response to the ongoing electric vehicle (EV) price competition, which has caused a decline in the company’s profit margin for multiple consecutive quarters. In 2023, the corporation achieved a remarkable milestone by delivering a total of 1.81 million automobiles, setting a new record. However, Tesla’s profit margins were negatively impacted as they constantly reduced prices in order to increase sales and gain a competitive advantage.

Tesla has notified its employees that around 14,000 workers, which accounts for over 10% of the global organization’s workforce, will be laid off. These job cuts will affect Tesla’s businesses in the United States, Europe, and China. The layoffs, which impacted staff in all departments and at various levels of seniority, were implemented to decrease expenses and enhance efficiency in anticipation of the company’s forthcoming expansion, as stated in an internal email from CEO Elon Musk that has been accessed.

High achievers also reduce
According to two anonymous sources who spoke, a significant number of the people that were let go were highly skilled and accomplished. A source expressed astonishment at the magnitude of skilled personnel laid off and observed that a significant percentage of those impacted were engaged in projects that had been deprioritized by Tesla. The source refrained from specifying the particular projects.

According to sources, certain divisions experienced layoffs that exceeded the 10% reduction mentioned in the companywide email. Management informed me that a reduction of 20% was made to their workforce.

“I experienced a 20% reduction in my team, including some highly skilled players,” they stated.

The shakeup is occurring concurrently with Musk’s ongoing efforts to steer the company towards the development of fully autonomous vehicles. Tesla has lately abandoned its ambitions to manufacture a more affordable electric vehicle (EV) with a starting price of approximately $25,000. Instead, they have decided to utilize the ongoing development of a foundational platform to support a purported robotaxi, which Elon Musk claims will be unveiled on August 8th.

According to his biographer, Walter Isaacson, Musk made an effort in the past to give priority to the project focused on developing vehicles specifically for robotaxi services. In 2022, he informed the employees of his desire for a “clean robotaxi” devoid of any steering wheel or pedals. Tesla’s lead designer, Franz von Holzhausen, and engineering VP, Lars Moravy, clandestinely spearheaded the development of the affordable electric vehicle (EV) project and eventually persuaded Elon Musk to pursue both options. However, recent reports indicate that Musk has now reversed his decision.

High-level executives depart
Two prominent executives, Drew Baglino, Tesla’s Senior Vice President of Powertrain and Energy, and Rohan Patel, Vice President of Public Policy and Business Development, have both departed from the business.

Patel informed me that he made the decision to resign from Tesla on Sunday evening due to significant general changes occurring within the firm. Patel, who had been routinely interacting with Tesla customers and enthusiasts on X in recent months, chose not to provide detailed information. In his message, he acknowledged that it would be more prudent for him to refrain from making conjectures. “Tesla will experience a significant increase in strength and resilience, and change will be beneficial,” he further stated.

Baglino informed me that he decided to depart from Tesla after a period of 18 years. “I am satisfied with the positive influence I have had; my leadership team is competent, and the energy businesses under my responsibility are performing well,” he expressed in a message.

“Baglino’s responsibility included overseeing powerdrives and new battery projects. According to Sandeep Rao, the head of research at Leverage Shares, a financial services company based in London, there is a perception that there is currently a lack of sustainable innovation. This perception may be one of the reasons why Baglino is leaving,” speculated Rao in an interview.

Baglino’s resignation occurred shortly after the departure of Tesla’s former CFO, Zachary Kirkhorn. In January, Musk expressed his desire on X, then known as Twitter, to obtain approximately 25% voting control of Tesla. His intention is to devote more attention to the firm itself than to his other ventures and contribute to Tesla’s leadership in the fields of AI and robots.

 

Continue Reading

Artificial Intelligence

The Holodeck from Star Trek has been replicated as a virtual training environment for advanced robots of the future

blank

Published

on

blank

Many Star Trek enthusiasts have pondered the possibilities of encountering a new species, swiftly teleporting out of an uncomfortable social situation, or zooming through space at warp speed (just be cautious not to exceed warp 10!). Many of the impressive technological advancements portrayed in the franchise are still confined to the realm of fiction. However, certain innovations inspired by Star Trek have indeed materialized in the real world. Thanks to a group of brilliant minds at the University of Pennsylvania, we can now proudly include a holodeck on that remarkable list.

Let me clarify something as we share your disappointment. We’re not referring to a futuristic world where humans can engage with characters in a holographic environment. That kind of technology is still a long way off. A team at Penn Engineering and their collaborators created the Holodeck system, which has the remarkable ability to create a wide variety of 3D environments. All you need to do is inquire.

“Language can be utilized to exert control,” stated Yue Yang, one of the co-creators. You have the ability to effortlessly describe any environments you desire and train the AI agents that inhabit them.

The holodeck system portrayed in Star Trek series such as The Next Generation and Voyager is a highly adaptable virtual environment capable of transforming simple verbal commands into fully simulated worlds. These types of environments, although smaller in scale, have significant applications in training robots.

Developing a virtual world can be quite time-consuming. “Creating these environments requires manual effort, with artists dedicating a significant amount of time to building just one,” Yang explained. To effectively train a robot to navigate real life, it is crucial to expose it to a diverse range of environments for testing purposes. Generative AI, which has gained significant popularity in recent months, appeared to be the obvious solution.

“AI systems such as ChatGPT undergo extensive training on an enormous amount of textual data, while image generators like Midjourney and DALLE are trained using a vast collection of images,” explained Chris Callison-Burch, an Associate Professor specializing in Computer and Information Science at the University of Pennsylvania.

The Holodeck utilizes a sophisticated language model known as a large language model (LLM). This powerful system, similar to the ones used in chatbots like ChatGPT, engages in a conversation with the user to determine the specific parameters of the desired environment. The system utilizes a vast digital library known as the Objaverse, which contains countless preexisting objects. It can effortlessly choose appropriate furnishings from this collection. Additionally, a layout design module ensures that the spatial arrangement of these objects is logical and coherent within the room.

blank

Practically speaking, if you inquire about the apartment of a cat owner, Holodeck will make sure that the final room is equipped with all the necessary furniture, including a cat tree.

The team conducted a comparison between Holodeck and a previous tool called ProcTHOR. They created 120 scenes and administered a blinded test to students, asking them to indicate their preferences. Holodeck clearly outperformed its competitor in every aspect. The system demonstrated its versatility by successfully generating a wide range of unique spaces, including science labs and wine cellars.

According to co-creator Assistant Professor Mark Yatskar, the ultimate test of Holodeck is its ability to assist robots in safely navigating unfamiliar environments.

Virtual training typically focuses on residential environments, but there are countless unfamiliar worlds that a robot may encounter and must learn to navigate. Utilizing Holodeck instead of the previous tool had a significant positive impact. For instance, a robot that underwent pre-training on 100 virtual music rooms created by Holodeck demonstrated a 30 percent success rate in locating a piano, compared to just 6 percent after training with ProcTHOR.

This holodeck has the potential to make a significant impact in the field of robotics, although it may not be suitable for running a Dixon Hill holonovel.

The study is scheduled to be presented at the 2024 IEEE/CVF Computer Vision and Pattern Recognition Conference. An unpublished paper that has not undergone peer review can be accessed through arXiv.

Continue Reading

Trending