Morrissey Technology

Loading

Tag Technology News

Gemini Live
News

Google Launches Gemini Live, AI That Can Be Chatted With

Morrissey Technology – Google officially launched Gemini Live, an AI assistant that can be chatted with at the Made by Google event on Thursday (14/8). Gemini Live was previously introduced at I/O 2024 last May. Now, this feature has been officially launched to become a chat friend for users. Unfortunately, not everyone can use this AI assistant. Google launched Gemini Live on Gemini Advanced, Google’s paid version of AI.

“Gemini Live will start rolling out tomorrow, it will be available to all Gemini Advanced users on all devices. So, the rollout may take a week or two to fully complete, but it will be available on all Android phones on Gemini Advanced,” said Amar Subramanya, VP of Engineering at Google in an exclusive Virtual Roundtable session on Gemini News, Wednesday (13/8).

“We’ll start with English on Gemini Advanced, with more languages ​​on the way,” he added.

Gemini Live represents a giant leap forward in AI assistants. The reason is, users can now chat with AI, instead of just asking questions. Previously, AI assistants could only interact in a question and answer format. In this session, Amar demonstrated Gemini Live by asking this AI to speak to those of us present at the virtual session.

“Hello to everyone there. I am very happy to be able to meet you all, even if only through my voice. I hope I can meet your expectations and you will enjoy learning about me,” said Gemini.

Amar interrupted Gemini’s words several times, and Gemini responded by stopping talking while listening to Amar’s words.

“This interaction is much more fluid, compared to taking turns talking. In my experience, I use it more as a trainer or as a brainstorming partner,” said Amar, recounting his experience of using Gemini Live for several weeks.

Amar said that currently Gemini Live can only input and output sound. However, other modalities such as visuals are being worked on by Google to provide this AI assistant. Google itself provides the option of 8 sound options in this initial version of Gemini Live. Furthermore, currently Android cellphone users can access Gemini for free. However, users who want to use Gemini Advanced must subscribe at a cost of IDR 309 thousand per month.

https://www.tomjonespottery.net/

Cisco Lays Off Thousands of Employees
News

Cisco Lays Off Thousands of Employees

Morrissey Technology – Cisco will lay off thousands of employees as part of its shift in business direction to more promising sectors, such as cybersecurity and AI. This is the second wave of layoffs carried out by Cisco, and it is estimated that the number of employees laid off will be almost the same, or perhaps even higher than before, namely when Cisco laid off 4,000 employees last February.

Reported by Reuters, citing an anonymous source, these layoffs will be officially announced at the same time as Cisco’s Q4 2024 financial report. Cisco shares immediately fell 1% after news of these layoffs was first announced, and Cisco shares have fallen 9% since the beginning of 2024. As of July 2023, Cisco had 84,900 employees, but this number has not been reduced by employees who were laid off in February 2024.

Cisco, a manufacturer of high-end routers and switches, said its business was struggling due to slumping demand and problems in the supply chain. This means they are diversifying their business, one of which is by acquiring the cybersecurity company Splunk for USD 28 billion in March 2024. This acquisition is expected to reduce Cisco’s dependence on network equipment sales revenue and increase their subscription business. Then Cisco is also trying to expand its AI business by incorporating AI into its products. They also target to get revenue of USD 1 billion from AI products by 2025.

Last June they injected funds into various AI startups amounting to USD 1 billion, for example Cohere, Mistral AI, and Scale AI. They also said they had made 20 AI-related acquisitions and investments over the past few years. Based on the monitoring site Layoffs.fyi, 126 thousand people have been laid off from 393 technology companies during 2024. The latest news comes from Intel, which laid off 15% of its employees in early August, or around 17,500 employees.

https://www.ayuntamientodegallegosdesolmiron.org/

Apple Mac Mini
News

Apple Mac Mini Will Get Even Mini

Morrissey Technology – Apple is said to be overhauling the design of the upcoming Mac Mini, and this will be the first major design change since 2010. In this overhaul, Apple will reduce the dimensions of the Mac Mini, to the point that it could possibly be called Apple’s smallest desktop computer. According to Bloomberg’s Mark Gurman, citing insider sources, the new Mac Mini will be similar in dimensions to the Apple TV, only slightly thicker.

Currently, Apple is testing a new Mac Mini design that has at least three USB-C ports and an HDMI, and the casing is still made of aluminum. Apart from design, the 2024 Mac Mini will also use the M4 chip, which is expected to be used gradually in the Mac product line, both desktops and laptops. The M4 was first released on the iPad Pro, and will later also be available on the Mac Mini, MacBook Pro, MacBook Air, iMac, and Mac Studio.

The last Mac Mini released by Apple is the 2023 release, which has two variants. One is with the M2 chip and the other is with the M2 Pro. This step will also be continued on the 2024 Mac Mini, which will use the M4 and M4 Pro chips, but the Pro variant will only be available after October 2024. Previously it was reported that the iPad Pro received a significant performance increase after being updated from the M2 chip to the M4.

When tested, the M4 chip with the 10-core variant recorded an average single-core score of 3,695 and an average multi-core score of 14,550 from the results of 10 benchmarks. Judging from the single-core score, the M4 chip is much faster than the MacBook Pro with the M3 Max chip. The M4 chip’s multi-core score also nearly matches the M2 Max chip. For comparison, here are the single-core and multi-core scores of all Apple M-series chips released in the last two years:

  • M4: 3.695/14.550
  • M3: 3.087/11.702
  • M3 Pro: 3.112/15.286
  • M3 Max: 3.128/20.957
  • M2: 2.540/9.360
  • M2 Pro: 2.651/14.295
  • M2 Max: 2.802/14.800

When compared to the iPad Pro which uses the M2 chip, the M4 version of the iPad Pro has 46% faster single-core performance and 55% faster multi-core performance.

https://www.bandeinrete.net/

Intel Panther Lake Processors
News

Good News from Intel: Panther Lake Processors Are Ready to Show Off

Morrissey Technology – Behind the various problems that have hit Intel, it turns out there is still good news from the company that makes the processor. The good news is regarding the next generation of processors codenamed Panther Lake and Clearwater Forest, which are made with the 18A process node. According to Intel SVP and GM Foundry Services Kevin O’Buckley, the Panther Lake processor has been successfully tested and booted into the operating system.

By successfully booting the operating system without additional configuration or modification, Panther Lake and Clearwater Forest clearly demonstrate the health of Intel 18A – Intel’s leading process technology, which is expected to return Intel to the ranks of leaders in process technologies by 2025.

Apart from that, Panther Lake DDR memory performance can operate at the targeted frequency. Meanwhile, Clearwater Forest, which is the archetype of future CPU and AI chips, combines RibbonFET, PowerVia, and Foveros Direct 3D for higher density and power handling.

Clearwater Forest is also a key product for Intel 3-T base-die technology. By leveraging Intel Foundry’s foundry system approach, these two products are expected to provide significant improvements in performance per watt, transistor density, and cell utilization.

“We are pioneering foundry system technologies for the AI ​​era and delivering a complete set of innovations that are important for the next generation of Intel products and our foundry customers. We are encouraged by our progress and working closely with customers to bring Intel 18A to market in 2025,” said O ‘Buckley.

In July, Intel released the 18A Process Design Kit (PDK) 1.0, a design tool that allows foundry customers to leverage the capabilities of the RibbonFET gate-all-around transistor architecture and PowerVia backside power delivery in their designs on the Intel 18A. Electronic design automation (EDA) and intellectual property (IP) partners update their offerings so customers can begin their final production designs.

This achievement shows that Intel Foundry is the first to successfully implement RibbonFET gate-all-around transistors and PowerVia backside power technology for foundry customers. Through the EDA ecosystem and IP tools and process flows, RibbonFET and PowerVia are innovative advances brought by Intel Foundry to all customers through Intel 18A.

Intel partners gained access to Intel 18A PDK 1.0 last month, EDA and IP partners updated their tools and design pipeline so external foundry customers can start designing their Intel 18A chips. This is an important milestone for Intel’s foundry business.

“Cadence’s strategic collaboration with Intel Foundry has helped accelerate our mutual customers’ innovation by providing access to industry-leading EDA solutions and IP optimized for Intel 18A,” said Tom Beckley, senior vice president and general manager, Custom IC & PCB Group at Cadence.

https://wvaaa.com/

https://www.htca.org/

Nvidia
News

Nvidia Caught Taking AI Data from Netflix and YouTube

Morrissey Technology – According to a report from 404 Media, Nvidia has used videos from YouTube, Netflix and other sources to build an AI model. Nvidia says its AI training methods are fully compliant with content and copyright laws. However, there is no definitive verdict on whether using an artist’s work to train an AI model is legal or not.

This particular model has not yet been released by Nvidia, but internally the AI ​​model is called Cosmos and will be the foundation model for advanced video that can power various products such as world generators and digital humans. 404 Media obtained Slack messages showing Nvidia employees using a YouTube video downloader to grab content while discussing the legal and ethical considerations of the practice.

“We respect the rights of all content creators and believe that our models and research efforts fully comply with the letter and spirit of copyright law,” an Nvidia spokesperson told 404.

“Copyright law protects certain expressions, but it does not protect facts, ideas, data, or information. Anyone is free to learn facts, ideas, data, or information from other sources and use them to create their own expressions. Fair use also protects the ability to use a creation for transformative purposes, such as model training,” he continued.

However, YouTube views this practice as a violation of its policies, while Netflix also says that scraping is against its terms of service. According to the leak, Nvidia downloaded 100,000 videos from YouTube in just two weeks and collected more than 38.5 million video URLs. These links include content creators such as Marques Brownlee and the Architectural Digest channel. The documents also show that Nvidia trained its models on a dataset called HD-VG-130M that contains 130 million YouTube videos and is explicitly for academic research only. This leak makes it clear that Nvidia’s work is for commercial gain. When an employee raised legal and ethical concerns, Ming-Yu Liu, Nvidia’s Vice President of Research and leader of the Cosmos project, told them that the decision to obtain data this way had been made at the top of the company.

Nvidia has grown into a trillion-dollar company thanks to its computer chips that are the foundation for the booming AI market. OpenAI, Microsoft, Meta, and Google are Nvidia customers and rely on its graphics processing units (GPUs). However, Nvidia does more than just hardware. Last week, Getty Images announced a deepening of its relationship with the company after releasing an updated AI image model based on the Nvidia Picasso model architecture.

https://www.lafayettecob.org/

AI Characters
News

Get to know AI Characters for Chatbots and How to use Them

Morrissey Technology – Artificial intelligence or Artificial Intelligence (AI) is currently becoming a trend among society. In fact, there are a number of AI sites that are popular and widely used, one of which is Character AI. Through Character AI, detikers can communicate with chatbots. However, the chatbot on this AI site is said to be very ‘smart’ because it is similar to talking to a human. Therefore, many people are curious about the chatbot feature in Character AI. Want to know how to use it? Read more in this article.

Understanding Character AI

Character AI is a chatbot application and website that can be communicated with and produce amazing responses. Through this AI chatbot, you can chat with a number of fictional characters, such as famous figures, celebrities and historians in a dialogue. Reported by the Elegant site, the chatbot output on Character AI is considered to be more similar to humans than other AI chatbots, for example ChatGPT. As a result, you can communicate with the chatbot like chatting directly with someone. In fact, users can chat with more than one bot at a time.

For those who don’t know, a chatbot is a computer program created to simulate conversations with humans, both in text and voice form. This system is designed with an interactive model that can interact with humans. Character AI was created by former Google AI developers, Noam Shazeer and Daniel De Freitas. Character AI was initially released in beta form in September 2022. Since its official launch, this AI chatbot has become one of the most popular besides ChatGPT.

How Character AI Works

Character AI is a chatbot that works using large language models (LLM). This AI chatbot is also supported by machine learning which focuses on conversation. During the training process, Character AI continuously reads large amounts of text, then learns to determine which words are likely to appear next in a sentence. The use of this sophisticated technology produces chatbots that are entertaining and look real, so it’s like talking to a real person.

The uniqueness of Character AI

Character AI has its own uniqueness compared to other AI chatbots. Apart from communicating like humans, Character AI has a number of characters that exist in the real world, such as Elon Musk and Taylor Swift. Of course, they are chatbots so they are not real people. We tried one of the Character AI features by communicating with Elon Musk’s character. As a result, this chatbot understands and understands the opinions expressed, especially when it comes to Tesla and SpaceX, which are companies owned by Elon.

Another unique thing about Character AI is that users can choose various chatbots according to the theme. There are around 20 chatbot themes available, ranging from helpers, famous people, games, VTubers, and game characters. Users can also create their own fictional characters and chat with them. In Character AI, you can also create characters according to their personalities and add avatars that can be created freely.

Disadvantages of Character AI

Even though it is called one of the smartest chatbots because it is similar to talking to humans, there are still a number of shortcomings of Character AI. One of the shortcomings of Character AI is when asked to create an image. Because the base is a text program, the chatbot in Character AI cannot create perfect images like an AI art generator. Apart from that, the chatbot also blocks any content that is deemed unsafe for use (NSFW or not safe for work).

Is Character AI Free?

Character AI can be used for free by accessing it on the character.ai site. You can also chat directly with as many chatbots as you wish. However, Character AI also provides a premium version called Character AI Plus (c.ai+). A number of benefits from Character AI Plus include:

  • Priority access
  • Faster chatbot response time
  • Early access to try new features
  • Access to the Character AI Plus community
  • Gets a badge profile.
  • For just US$ 9.99 or around Rp. 160 thousand per month, you can enjoy the Character AI Plus service with satisfaction.

How to Use Character AI

Want to try chatting with a chatbot in Character AI? Here’s how to use it:

  • Open the character.ai site in a browser or download the Character AI application on a smartphone
  • If you don’t have an account, please log in using a Gmail or Apple account
  • After logging in, select one of the fictional characters according to the available theme. Or you can also search for it in the search bar column
  • If so, click on the character account
  • Ask the character a question or just say hello
  • After that, you can ask questions or chat with the chatbot like chatting with a human.

https://www.symposiumgeorgia.org/

https://www.human-body.org/

https://webua.org/

https://www.wartasundaonline.com/

https://www.the-uelci.org/

https://www.iclc2001.org/

Study Reveals Hackers Can Steal Data Via HDMI Cables
News

Study Reveals Hackers Can Steal Data Via HDMI Cables

Morrissey Technology – A recent study by a research team at the Universidad de la Republica Montevideo, Uruguay, revealed that hackers can use artificial intelligence (AI) and HDMI cables to steal data on their victims’ computers. Read the explanation. The research, entitled ‘Deep-TEMPEST: Using Deep Learning to Eavesdrop on HDMI from its Unintended Electromagnetic Emanations’, was published on the arXiv preprint server. The research was conducted by Santiago Fernandez Emilio Martinez, Gabriel Vareal, and Pablo Muse Federico Larroca.

HDMI cables are usually used to transfer the screen display from a laptop or PC to a projector to have a larger display. This cable apparently produces electromagnetic radiation that can be captured by hackers. In this study, researchers found that hackers also used AI technology in hacking practices via HDMI cables.

The study involved capturing electromagnetic radiation emitted from a computer’s HDMI cable. They then trained the AI ​​system by providing screen samples associated with radiation signals traveling through the computer’s HDMI cable. As time went by, these systems gradually became better at interpreting text displayed on computer screens.

Test results show this system is able to reconstruct text from a computer screen randomly with 70 percent accuracy. They also note these systems may be good enough to steal passwords, sensitive data, or in some cases, encrypted communications. The researchers found that they could improve their results by using text recognition software on the text after it had been parsed.

Researchers say hackers may have conducted similar research. This means such screen hacks may have targeted a number of victims. According to them, all that is needed is hardware capable of capturing the electromagnetic radiation emitted by HDMI cables placed close to buildings, such as in the back seat of a car. The team further suggests that most people are not at risk of such an attack, due to the expertise required by this technique. Instead, they suggest that it is more likely that governments or corporate entities will be targeted.

https://www.balaitradisionaldenpasar.org/

Beware of AI Application Ads on Facebook, Accounts Can Be Stolen
News

Beware of AI Application Ads on Facebook, Accounts Can Be Stolen

Morrissey Technology – The trend of using artificial intelligence (AI) has been spreading in recent times. Not surprisingly, many technology companies promote their AI products through advertisements on social media. However, users also need to be careful and not carelessly click on AI application advertisements on social media, especially Facebook. The reason is, a fake ad on Facebook targets users looking for an AI image editing tool and steals their credentials by tricking them into installing a fake application that imitates the original software.

According to Bleeping Computer, hackers are exploiting the popularity of AI-based image creation tools by creating malicious websites that closely resemble the original service and tricking potential victims into infecting themselves with information-stealing malware. This was discovered by Trend Micro researchers who analyzed the ad.

The attack began with a phishing message sent to the Facebook page owner or admin, which would redirect them to a fake account protection page designed to trick them into providing login information. After stealing their credentials, the perpetrators hijacked the accounts, took control of their pages, published malicious social media posts, and promoted them through paid advertising.

“We discovered a malvertising campaign involving threat actors stealing social media pages (often related to photography), changing their names to appear connected to a popular AI photo editor,” said Trend Micro researcher Jaromir Horejsi.

“The threat actor then created a malicious post with a link to a fake website that was made to resemble a legitimate photo editor website. To increase traffic, the perpetrator then boosted the malicious post through paid advertising,” he continued.

Facebook users who click on the URL in the malicious ad will be sent to a fake web page masquerading as genuine AI photo editing and creation software, and will then be asked to download and install a software package. Instead of installing AI image editing software, victims installed an ITarian remote desktop tool that was configured to launch a downloader that automatically deployed the Lumma Stealer malware.

This malware then silently infiltrates their systems, allowing hackers to collect and compromise sensitive information such as credentials, cryptocurrency wallet files, browser data, and password manager databases. This data is then sold to other cybercriminals or used by attackers to break into victims’ online accounts, steal their money, and promote further fraud.

“Users should enable multi-factor authentication (MFA) on all social media accounts to add an extra layer of protection against unauthorized access,” advises Horejsi.

“Organizations must educate their employees about the dangers of phishing attacks and how to recognize suspicious messages and links. Users should always verify the legitimacy of links, especially those that ask for personal information or login credentials,” he continued.

https://www.karangasem-bangkit.org/

malware
News

Horrified, a New Virus Appears That Drains Your Account and Deletes Your Cellphone Data

Morrissey Technology – A malware or malicious software on Android called BingoMod was found to be able to drain accounts and delete cellphone data. This malware was found to be promoted via text messages, and masquerades as a legitimate mobile security tool. This malware can steal up to 15,000 Euros or IDR 262 million per transaction. According to the researchers who analyzed it, BingoMod is currently under active development. They say malware authors are focusing on adding code obfuscation and various evasion mechanisms to reduce detection rates.

Researchers at Cleafy, an online fraud management and prevention solution, reported BleepingComputer, found BingoMod distributed in smishing (SMS phishing) campaigns. They use various names that usually indicate mobile security tools such as APP Protection, Antivirus Cleanup, Chrome Update, InfoWeb, SicurezzaWeb, WebSecurity, WebsInfo, WebInfo, and APKAppScudo. In one example, the malware used the icon for the free AVG AntiVirus & Security tool available on Google Play. During the installation process, the malware requests permission to use the Accessibility Service, which provides advanced features that allow broad control over the device.

Once active, BingoMod steals any login credentials, takes screenshots, and reads SMS messages. To perform on device fraud (ODF), the malware creates a socket-based channel to receive commands and an HTTP-based channel to send screenshot feeds, thereby enabling near real-time remote operations. ODF is a common technique used to initiate criminal transactions from a victim’s device, circumventing standard anti-fraud systems that rely on identity verification and authentication. Cleafy researchers explain the Virtual Network Computing (VNC) activity abusing Android’s Media Projection API to obtain screen content in real-time.

Once received, the content is converted into a suitable format and transmitted via HTTP to the cybercriminal or threat actor’s infrastructure. One of the features of this activity is that it can leverage Accessibility Services to impersonate users and enable screen-casting requests, which are exposed by the Media Projection API. Commands that a remote operator can send to BingoMod include clicking on certain areas, writing text on certain input elements, and launching applications. This malware also enables manual overlay attacks via fake notifications initiated by cybercriminals. Additionally, devices infected with BingoMod can also be used to further spread malware via SMS.

https://www.gvardeysk.org/

TikTok AI Cloud
News

TikTok Revealed to Be a Big Customer of Microsoft’s AI Cloud, What Can It Do?

Morrissey Technology – It was revealed that TikTok is one of the largest consumers of AI Cloud Computing managed by Microsoft. What can video-based social media do with AI? Quoting from The Verge, TikTok paid at least US$20 million (Rp. 322.77 billion) per month to Microsoft to access the OpenAI model last March. That amount is almost equivalent to a quarter of Microsoft’s Cloud division’s revenue. Microsoft’s AI cloud business, according to The Information, is currently achieving annual revenue of around US$1 billion (Rp. 16.14 trillion). Microsoft is also investing multibillion US dollars to make OpenAI its proprietary cloud, and has also poured hundreds of billions of dollars into building the supercomputer that powers ChatGPT.

On the one hand, a study reports that TikTok may not need Microsoft’s cloud AI capabilities if it develops its own Large Language Model (LLM). Last year, the company that houses TikTok, ByteDance, secretly used OpenAI technology to create its own LLM. That practice is generally considered reckless in the world of AI. It is also a direct violation of OpenAI’s terms of service, which state that its model output cannot be used ‘to develop any artificial intelligence models that compete with our products and services’. Microsoft, from which ByteDance buys its OpenAI access, has a similar policy.

Following the report, OpenAI suspended ByteDance’s account to investigate potential violations of its developer license. At the time, ByteDance said it was using the technology ‘to a very limited extent’ to help build its own language models.

AI search feature on TikTok

In May, TikTok was reportedly testing an AI Search Highlights feature on its search page. Quoting The Verge, this AI-based feature will summarize the contents of various videos and highlight them. On the AI ​​disbursement highlights feature page, TikTok explains that its material is supported by ChatGPT made by OpenAI. Meanwhile, at the bottom of the AI ​​search results, TikTok will display videos that the algorithm system considers relevant.

Al Smart Search TikTok provides ChatGPT answers to search queries. TikTok is ultimately trying to capitalize on some of the habits its users – especially the younger generation – have developed on the app. Many young TikTokers treat TikTok like a search engine, choosing to look for restaurant or product recommendations on the app rather than using platforms like Google Search. And like Google, TikTok now seems to be leaning towards AI results by incorporating content from top content creators.

https://www.santacroyadetera.org/