Since the start of this year, Microsoft, eBay and PayPal have axed thousands of jobs in a pivot to artificial intelligence as companies revise their resources.

As reported by the Financial Times, tech companies have cut around 34,000 jobs in a matter of weeks as the focus turns to generative AI.

In addition to the heavyweight names mentioned above, Snap has also shed a significant number of roles as part of an industry-wide cut of 138 firms that have let staff go. These figures are significant but not quite at the same level as one year previous when Microsoft, Meta, and Amazon culled their numbers.

Layoffs.fyi — indicated 263,000 roles were lost in the tech sector during 2023 as analysts have indicated the latest round of cuts is a sign of the pivot to AI, while some companies have considered productivity relating to their overall output.

Brent Hill, an analyst at the renowned financial institution Jefferies, opined companies have assessed their workforces with the conclusion that “we’ve got a bunch of dead wood, and if we had a leaner organization, we can do more.”

“The lay-offs are going to continue, and it may get worse. It’s become contagious,” added Hill.

Who is next?

Last month, Amazon-owned Twitch was hit with the blow of around 500 employees being axed due to financial pressures, a further indicator of the cost-cutting and adjusting across the industry. Even the ‘Big Five’ are impacted as part of an extensive list of employers who have cut their cloth over the last few weeks.

Enterprise software company SAP confirmed a “company-wide transformation” in January, including the loss of around 8,000 jobs as it turns further to AI while reskilling staff as much as possible.

Autumn Mitchell, a QA tester at Microsoft subsidiary ZeniMax, commented on the current uncertainty, “Anybody working in tech or games right now is worried about lay-offs to some degree, either for themselves or someone they know.”

“You see one company announce lay-offs and think, Here we go, who’s it going to be next week?” Mitchell said.

Image: PVMS/Pexels.

The post Microsoft, eBay, PayPal among tech giants 34,000 job cuts due to AI appeared first on ReadWrite.

Nearly every industry is being transformed by AI’s familiar yet disruptive technology, including the real estate sector. Its impact has become more compelling as institutions realize its generative capabilities.

The real estate sector isn’t too far ahead in using AI as an aid for residents, builders, mortgage brokers, agents, and contractors. Going beyond ChatGPT, other platforms can help streamline other more complicated areas in real estate. It’s hard to tell what AI will do, but it’s already changing the real estate sector.

AI is Impacting Real Estate. Here’s How.

1. Generating and validating real estate leads

It remains a challenge for real estate brokerages to gather contacts that match their ideal client profiles. The cost and time needed to generate high-quality leads have always been a critical issue, but with AI platforms, easing up the process is possible.

Using predictive analytics, real estate businesses can seek out leads with updated addresses and contact details. AI can also help clean up entire databases to remove invalid contacts and redundancies. This helps increase the success rate of marketing campaigns and avoids DNC policy violations.

2. Running accurate property valuations

For property owners looking to put their homes up for sale, selling at the right price has always been a major hindrance. There are numerous factors to consider such as the age of the property and the median selling price within the neighborhood.

This won’t be ideal for homeowners who, according to a home cash buying company, have to deal with too many repairs and avoid foreclosure. AI streamlines the task by using algorithms to create accurate valuations and propose sweet spot asking prices.

3. Ease up property management

For landlords and investors who own entire condo buildings or townhouses, keeping track of tenant satisfaction and property health has always been a matter of guesswork. This is expected to change with the inclusion of AI in property management processes. Using the right software, property owners can accurately measure rent collections, ROI, and maintenance expenses.

Not only that, AI can help reduce the paperwork needed for compliance and handle the more intricate parts of tenant screening and tenant communications. At the moment, there are software packages on the market that are specifically meant for property management, offering owners and landlords a great way to stay profitable.

4. Come up with an accurate market analysis

Market research serves as the guiding force for realtors. How buyers and sellers behave provides insight into the overall health and the availability of opportunities on the market. Rather than conduct surveys, real estate professionals can use AI software to analyze and predict market trends. That way, they will be able to make informed decisions in crafting marketing campaigns and business strategies.

Whether you like it or not, AI is here to lessen the grunt work in the real estate sector. It will make life easier for buyers, sellers, and investors. It’s only a matter of using the right software and understanding what else this technology is capable of in the near future.

The post How AI Is Transforming the Real Estate Sector appeared first on ReadWrite.

Google has extended the availability of its Gemini artificial intelligence (AI) assistant and chatbot to several more territories and in various languages.

Last week, the Alphabet company confirmed the rebranding of Bard to Gemini, which is also the name of its flagship large language model (LLM) AI. The Gemini assistant was rolled out for Android and iOS, but only to users in the United States, in the English language. It was expected that availability would be rolled out to other countries soon after, which has now been made official.

As reported by Sam Mobile, Android users across Asia have confirmed access to Gemini on their devices but there is no sign of the feature in Europe just yet. This rollout could be impacted by regulatory issues in the European Union but given the size of its consumer base there, efforts are surely ongoing to fix any issues.

On the Android devices that Gemini accesses, consumers can access it from its app, as well as by long pressing the power button, swiping diagonally from the lower corners of the screen, or saying “Hey Google.” It first appears on top of other apps, and you can type your question using the keyboard or by using your voice.

Advanced UI

Jack Krawczyk, a lead developer on Gemini, intimated the program would soon be arriving in Africa, Asia-Pacific, Latin America, as well as in North America, as Google works to make its AI assistant available to the world’s 2.5 billion Android users.

Gemini is powered by the LLM of the same name, which has a far greater capacity for understanding communications such as natural language and conversations. It is expected to fully replace Google Assistant in a matter of time given its advanced user interface and capability.

Gemini seamlessly integrates with Google services including Gmail, Google Docs, Google Drive, YouTube and more, allowing users to make prompts from their account on the apps.

Image: Google

The post Google Gemini AI now in more countries for Android users appeared first on ReadWrite.

Three students share the grand prize for using artificial intelligence (AI) to translate Pompeii scrolls found by an eighteenth-century Italian farmer.

The $700,000 Vesuvius Challenge grand prize was awarded by an international team of papyrologists (ancient paper specialists) to students Luke Farritor from the U.S., Youssef Nader from Egypt, and Julian Schilliger from Switzerland.

The students would successfully complete the task of being the first team to recover four passages, with a total of one-hundred-and-forty characters of the archaic writings.

The work the students carried out sent waves across the world and did not escape the attention of arguably the world’s richest man, Elon Musk, who posted on his social media platform that he would back more translations of antique texts;

Musk Foundation will support this

— Elon Musk (@elonmusk) February 7, 2024

The scroll of Herculaneum

The scroll centered around the AI challenge was named after the place where it was discovered, the small Roman town of Herculaneum, which was destroyed in the catastrophe of the Vesuvius eruption.

The texts stayed trapped in carbon for 2,000 years under the mud, ash, and slag that amazingly completely covered the villa of Julius Caesar’s father-in-law.

Many attempts were made to open them, but most destroyed the fragile contents, with a monk taking the time to painstakingly unwrap some of them, revealing ancient Greek text.

Six hundred would remain undisturbed, according to the grand prize’s website, and would lead to the young students getting their chance to unlock this historic time-trapped text.

They would follow in the footsteps of Dr. Brent Seales and his research team from the University of Kentucky, who used X-ray tomography and computer vision to read the scrolls without cracking the contents open. Their work paved the path to the historic work conducted by the students.

Youssef Nadar would post on his personal site “I am eternally grateful to be a part of such an incredible community, and having the chance to contribute to something so amazing.”

He continued, “This work was built on the shoulders of giants, first and foremost Prof. Seales and his team (I pretty much had Stephen’s Thesis open at all times), many amazing papers by FAIR and DeepMind, and so many open contributions from the community.”

Nebraska’s own Luke Farritor would be the first human to lay eyes on the contents and the winner of the first text prize in 2023. You can view his initial submission via Google Drive.

The technology and historical world will be keeping a close eye on the work of the young students and where, in time, their translation journey will take them.

Image Credit: Photo by Brent Keane; Pexels.

The post Students win $700,000 grand prize using AI to unearth Pompeii scroll’s secrets appeared first on ReadWrite.

1X Technologies has confirmed a breakthrough in teaching skills to its human-like (android) robots.

The company announced that its robot named EVE can now take more actions thanks to special neural networks, which allow the bots to learn by watching videos of people doing activities. From this, EVE learns to copy the activities on its own, without needing complex programming.

The Norway-based humanoid robotics firm, backed by OpenAI, designs androids to “work alongside people, to meet the world’s labor demands” but it assures of their safety credentials by testing every android in real-world scenarios before they’re deployed.

The post 1X robotics company showcases its androids driven by neural networks appeared first on ReadWrite.

Disney has introduced “Disney’s Magic Words,” an AI tool aimed at revolutionizing advertising on its streaming platforms, Disney+ and Hulu, according to a recent Reuters report. This new technology employs AI and machine learning to analyze scenes within Disney’s extensive content library, enabling advertisers to match their ads with the mood and context of specific scenes.

By identifying and tagging scenes with metadata that includes information on content, brands, imagery, and mood, “Disney’s Magic Words” allows for a more precise form of contextual advertising. This approach moves beyond traditional demographic targeting, offering advertisers the opportunity to personalize their messaging based on the emotional and thematic resonance of the content viewers are watching.

Geoffrey Calabrese, Chief Investment Officer at Omnicom Media Group, one of the six global advertising firms participating in the early beta test of the product, praised the tool for its potential to connect advertisers with the emotions of consumers on a more granular level. Other beta partners include Dentsu, GroupM, Horizon Media, IPG Mediabrands, and Publicis Media.

The technology was showcased last month at the Consumer Electronics Show in Las Vegas, where Rita Ferro, Disney’s global head of ad sales, emphasized its ability to enhance the impact of advertisements by aligning them with viewer experiences.

The initiative comes as advertisers increasingly shift their focus from traditional TV to streaming services, in line with changing viewer habits. Despite a nearly 3% decline in advertising revenue to $3.35 billion in the fiscal first quarter of 2024, Disney remains optimistic about the future of streaming advertisements. CEO Bob Iger reported that the ad-supported version of Disney+ attracted over 1,000 advertisers in the first quarter, a significant increase since its launch.

Joe Earley, President of Disney’s direct-to-consumer business, highlighted the immediate success of Disney+ in utilizing advanced ad technology, a capability that has been refined over years, particularly with Hulu’s ad-supported service in 2008. According to Earley, Disney+ was able to leverage this technology effectively from the start.

The post Disney unveils AI-powered ad tool for tailored streaming commercials appeared first on ReadWrite.

As if snatched straight from the world of Pokémon itself, an inventor has created an actual, working Pokédex using generative AI.

YouTuber abe’s projects has made a seeing, talking Pokédex using a 3D printer, ChatGPT, and some electronic components. It’s extremely cool and highly nostalgia-inducing for Pokémon fans of the 90’s.

A Pokédex is a device from the Pokémon anime and games that the protagonist uses to scan and identify the Pokémon they encounter. It then provides information about that Pokémon back to the user.

Abe recreated the device’s distinctive red case using a 3D printer, switching to beige for the bezels around the screen and buttons. Abe chose to use a black and white screen for his Pokédex despite the anime featuring a color screen because he wanted to call back to a toy from the ’90s which also had a black and white screen.

Abe incorporated a small camera in the front of the device for scanning Pokémon. It then connects with ChatGPT and the Pokémon API to determine what creature it’s seeing. Abe used a tool called PlayHT to modify the AI-generated voice to make it have the sound and cadence of the Pokédex in the anime for an even higher nostalgia hit.

Abe overcame many bugs in his quest to recreate the Pokédex

In the YouTube video, Abe outlined some of the bugs he had to overcome to get the device working. Early on he mentioned how fragile the miniature display screens are, having already broken one before completing the project. He also faced multiple software bugs and crashes, but his ingenuity allowed him to overcome them.

This included a strange ticking sound in the audio. Abe couldn’t identify how or why the sound was being added to his audio output but he was able to find the sound signature and manually code in a fix.

If you’re a fan of Pokémon, cool personal electronics projects, or both, the video is very much worth watching.

Featured image credit: abe’s projects YouTube video

The post A YouTuber has made a working Pokédex with AI appeared first on ReadWrite.

The CEO of OpenAI, Sam Altman, is seeking a $7 trillion investment to overhaul the world’s semiconductor industry, reports The Wall Street Journal.

Semiconductor chips are a hugely limiting factor in the world of artificial intelligence (AI) development due to the extremely high computing power needed to train AI models. With this investment, Altman is looking to negate a factor that significantly limits OpenAI’s growth potential.

Despite huge investment and strategic focus from the US government, the semiconductor chip industry is complex. Due to its strategic value and the importance of chips to AI development, it is a highly regulated industry and the US has tight controls on the export of chips. Altman has met with Commerce Secretary Gina Raimondo to discuss his ambitious plan.

“OpenAI has had productive discussions about increasing global infrastructure and supply chains for chips, energy, and data centers—which are crucial for AI and other industries that rely on them,” said a spokesperson for OpenAI to the WSJ. “We will continue to keep the U.S. government informed given the importance to national priorities, and look forward to sharing more details at a later date.”

However, fundraising the $5 – 7 trillion dollars required will be no easy feat. As the WSJ reports, the size of the global semiconductor chip market was $527 billion last year and isn’t expected to reach $1 trillion until 2030. The investment sought dwarfs even the market for chip manufacturing equipment, which was approximately $100 billion last year.

Sam Altman is looking for an amount greater than most country’s GDP

Let’s just stop for a moment and put this figure in context because $7 trillion is an extravagantly large amount of money. The GDP of the United Kingdom is $3.1 trillion, less than half of what Alman is seeking. The combined value of Microsoft and Apple, the USA’s highest-valued businesses, is just over $6 trillion.

The tech entrepreneur has met with representatives from the government of the United Arab Emirates and businesspeople from Taiwan to try and advance his plans. Despite Middle Eastern interest, Altman wants to focus his chip plants in the US, in part due to the huge investment coming from the Biden administration into the chip industry.

OpenAI is partnered with Microsoft and has their full support. Altman has discussed his plans with Satya Nadella, Microsoft CEO, and Kevin Scott, its CTO, and is proceeding with their buy-in.

Nothing has yet been confirmed regarding this enormous plan, but Altman continues to meet with powerful people to push it forward.

Featured image credit: Midjourney

The post OpenAI founder Sam Altman is ‘seeking $7 trillion investment’ appeared first on ReadWrite.

OpenAI is said to be working on a “supersmart” assistant to run a computer for its user.

With the development of a ChatGPT bot to effectively assume autonomy over a device, this is one of two agents in the pipeline, according to a report from Business Insider.

The leading artificial intelligence (AI) research company appears to have grand ambitions to deliver an ultimate assistant for the present era, which would result in the software being able to fully complete tasks as prompted. It will focus on text typing functions, as well as cursor movement, and working with different apps.

OpenAI CEO Sam Altman reportedly wants to deliver ChatGPT as a “supersmart personal assistant for work,” but the path ahead could lead to issues with Microsoft, its partner firm. The ‘Big Five’ tech giant (as well as Google) has already launched its own AI workplace assistant but these programs are said to be in their infancy.

The report suggests growing anticipation for the development from OpenAI, with work commenced on the project over a year ago. Talk of a game-changer will raise expectations but the end product is not in sight yet.

Matter of time

Generative AI has been lauded for significantly boosting workers’ productivity with some tasks, with the market for AI agents opening up in recent times. This growth area has been talked up by industry experts, including the co-founder of Google’s AI research subsidiary, DeepMind.

Mustafa Suleyman stressed that everyone could potentially have a personal assistant bot for themselves, a “chief of staff”, within the next five years in a conversation with CNBC on the rapid advances of AI.

“It will be able to reason over your day, help you prioritize your time, help you invent, be much more creative,” he said, adding “It will be a research assistant, but it will also be a coach and companion.”

Featured image: Dall-E

The post How OpenAI plans to make ChatGPT a ‘supersmart’ assistant using your computer appeared first on ReadWrite.

Nvidia, the leading global supplier of AI chips, is venturing into a new business domain by creating a dedicated unit for designing custom chips for cloud computing companies and other sectors, including advanced AI processors, according to sources cited in a recent Reuters report. This strategic move aims to tap into the burgeoning market for tailor-made AI chips and safeguard Nvidia’s dominance against the increasing competition seeking alternatives to its products.

Nvidia currently boasts an impressive 80% market share in the high-end AI chip sector. This dominance has significantly contributed to its market value, which has surged by 40% this year to $1.73 trillion, following a more than threefold increase in 2023. Major tech giants such as OpenAI, Microsoft, Alphabet, and Meta Platforms, all avid consumers of Nvidia’s H100 and A100 chips, are now exploring the development of their own custom chips to optimize energy usage and reduce costs.

According to the Reuters reporting, Nvidia’s initiative to assist these companies in developing bespoke AI chips could potentially divert business from competitors like Broadcom and Marvell Technology. The company’s custom chip unit, led by Dina McKinney, a former executive at Advanced Micro Devices and Marvell, focuses on making Nvidia’s technology accessible across various industries, including cloud, 5G wireless, video gaming, and automotive sectors.

The demand for custom chips in data centers is projected to reach $10 billion this year and is expected to double by 2025. The broader custom chip market was valued at approximately $30 billion in 2023, representing about 5% of the global annual chip sales. Nvidia’s entry into this market poses a significant challenge to Broadcom and Marvell, which currently lead in custom silicon design for data centers.

Nvidia’s broader ambitions may extend beyond AI and data centers. The company is reportedly in discussions with telecom infrastructure giant Ericsson for a wireless chip incorporating Nvidia’s GPU technology. Additionally, sources told Reuters that Nvidia aims to penetrate the automotive and video game markets, with plans to target next-generation consoles from Xbox and Sony, as well as a new version of Nintendo’s Switch console expected to feature a custom Nvidia design.

The post Nvidia to design tailor-made AI chips for cloud computing firms appeared first on ReadWrite.

With concerns over the impact of AI on the democratic process in the lead up to the 2024 United States presidential election, Midjourney has intimated it could ban political images entirely.

From Donald Trump campaigning with Donald Duck to Joe Biden being accompanied by Beyonce, anything is possible with artificial intelligence in terms of creating imagery and videos to blur the lines but worse than that is the fear of political disinformation – deliberately misleading – which has the potential to negatively influence public opinion in a meaningful way.

Generative AI program Midjourney, as well as other platforms, is cognisant of this issue with a decision to prohibit political imagery over the next 12 months said to be close.

Currently, as you can see from the headline picture here made on Midjournery, there is no restriction on creating images of the two men.

As reported by Bloomberg, Midjourney CEO David Holz opined in a Discord session, “I don’t know how much I care about political speech for the next year for our platform. I know it’s fun to make Trump pictures, I make Trump pictures,” he said.

“Trump is aesthetically really interesting. However, probably better to just not – better to pull out a little bit during this election. We’ll see,” added Holz.

Concern and control

Last March, Midjourney-created images of Trump being arrested went viral, as did pictures of the Pope depicted in a lavish Balenciaga jacket so it is easy to see where the concern of deepfakes comes from. The San Francisco-based company ended free trials soon after those pictures emerged.

Other tech firms are taking action too. OpenAI CEO Sam Altman commented at an event last week, “The thing that I’m most concerned about is that with new capabilities with AI, there will be better deepfakes than in 2020,” whilst Meta has confirmed it will be labeling more posts creating using AI in a direct move to target political interference.

It is a sensitive situation, requiring diligent action as 15 states prepare to go to the polls on Super Tuesday.

Image: Midjourney

The post Midjourney mulls Biden and Trump picture ban ahead of election appeared first on ReadWrite.

Researchers have developed the first-ever artificial intelligence (AI) child, Tong Tong (or Little Girl in English). It has the emotional intelligence and behavior of a three to four-year-old.

It has been developed by the Beijing Institute for General Artificial Intelligence (BIGAI) and according to the researchers, it can develop continually and improve its skills, knowledge, and values.

But don’t expect an army of robot toddlers to enter homes anytime soon. Tong Tong is virtual. She exists and learns in a virtual environment.

Unveiled at the Frontiers of General Artificial Intelligence Technology Exhibition in Beijing on 28-29 January, the AI can assign herself tasks based on how she has been programmed. For example, if she has been programmed to be tidy, she’ll clean up spills and straighten crooked picture frames. If she can’t reach the frame in her virtual world, she might get a stool to stand on to solve the problem.

𝐖𝐨𝐫𝐥𝐝’𝐬 𝐟𝐢𝐫𝐬𝐭 𝐀𝐈 𝐜𝐡𝐢𝐥𝐝

Tong Tong is a virtual AI entity that can assign herself tasks, learn autonomously, and explore her environment.

She can display behavior and abilities similar to those of a three or four-year-old child… pic.twitter.com/SoZpKLegwY

— CodingNerds COG (@CodingnerdsCog) February 6, 2024

This is a big difference between common large language model (LLM) AI, which cannot act autonomously, and general AI, which can. The developers also said that Tong Tong has her own emotions and can experience joy, anger, and sadness. Currently, emotion is not within the capabilities of AI so we should treat this claim with a healthy dose of skepticism.

According to the South China Morning Post, BIGAI stated in a video “Tong Tong possesses a mind and strives to understand the common sense taught by humans. She discerns right from wrong, expresses her attitudes in various situations, and has the power to shape the future.” Understanding right from wrong is also something current AI technology is unable to achieve.

On Tuesday (Feb.6) Readwrite reported how when given wargame simulations generative AI tools often chose violence, including launching nukes!

Of course, it has been seen time and again that AI might not always behave as expected with researchers recently encountering an AI deliberately rejecting instructions. Luckily, Tong Tong is only the size of a toddler and should be easy to stop if she gets too recalcitrant.

Tong Tong is the brainchild of Zhu Songchun, director of BIGAI. He stated, “To advance towards general artificial intelligence, we must create entities that can comprehend the real world and possess a wide range of skills.”

Zhu lived and worked in the USA for 28 years and was a professor at UCLA until he moved to China and established BIGAI. He is a world-class AI scholar with several awards and accolades to his name.

What is the Tong Test?

The Tong Test is a new test developed by Zhu and his team at BIGAI and is designed to test AI’s communication and intelligence across five dimensions – vision, language, cognition, motion, and learning.

On their website, BIGAI says, “With nearly 100 specialized tasks and more than 50 general tasks, the Tong Test offers a complete testing regime for the development of general artificial intelligence.” The statement on the website included a quote from Zhu: “For general AI to integrate seamlessly into human environments, it must learn and execute tasks in complex settings, driven by values and an understanding of causality. This is why we proposed the Tong Test, a new direction for testing general AI, focusing on practical abilities and values.”

Featured image: BIGAI

The post Meet Tong Tong, the slightly terrifying AI toddler appeared first on ReadWrite.

The Federal Communications Commission (FCC) has outlawed robocalls containing voices generated by artificial intelligence.

The ruling, which was unanimous, was made on Thursday, February 8. It gives the state power to prosecute “bad actors” under the Telephone Consumer Protection Act of 1991, a statute that prohibits junk robocalls with pre-recorded messages.

Under this consumer protection law, telemarketers are prohibited from using automated dialers or artificial voice messages in calls to cell phones. They also cannot make such calls to landlines without prior written consent from the call recipient.

Under this new ruling, AI-generated voices in robocalls are classed as “artificial” and enforceable in the same way.  Those who break the law can face steep fines of up to $23,000 per call.

Robocalls in politics

This new law comes after a recent scam where a ‘deepfake’ of Joe Biden’s voice encouraged New Hampshire residents not to vote in the recent primaries.

“Voting this Tuesday only enables the Republicans in their quest to elect Donald Trump again,” the voice said. “Your vote makes a difference in November, not this Tuesday.” The voice also adopted Biden’s much-used phrase: “What a lot of malarkey.”

It’s estimated that 5,000-25,000 calls were placed during this period, with the authorities later tracing the call to two companies in Texas.

“You know, it’s not like maybe it used to be where it costs thousands of dollars to create a deep fake. We’re talking about something that someone can do on their phone in five minutes time. It’s a really big problem, and it’s something we all need to be paying attention to,” New Hampshire state representative Angela Brennan told CBS.

Featured Image Credit: Photo by Alex Andrews; Pexels

The post Robocalls made with AI-generated voices are now illegal appeared first on ReadWrite.

Commerce Secretary Gina Raimondo announced the U.S. AI Safety Institute Consortium (AISIC). Raimondo said in a statement to Reuters, “The U.S. government has a significant role to play in setting the standards and developing the tools we need to mitigate the risks and harness the immense potential of artificial intelligence.”

The consortium members

Reuters published the list of consortium members, which includes BP (BP.L),  Cisco Systems (CSCO.O), IBM (IBM.N), Hewlett Packard (HPE.N), Northop Grumman (NOC.N), Mastercard (MA.N), Qualcomm (QCOM.O), Visa (V.N), and major academic institutions and government agencies, that will be housed under the U.S. AI Safety Institute (USAISI).

This group prioritizes the actions and guidelines listed in President Biden’s executive order:  “including developing guidelines for red-teaming (meaning identify new risks), capability evaluations, risk management, safety and security, and watermarking synthetic content.”

The executive order from U.S. President Joe Biden

Additionally, the Oct 30, 2023 executive order from U.S. President Joe Biden said that he “is seeking to reduce the risks that AI poses to consumers, workers, minority groups, and national security” with a new executive order. As per the Defense Production Act, creators of AI systems that endanger the national security, economics, health, or safety of the United States must notify the government of the United States of the findings of their safety texts before their public release.

In addition, agencies are instructed to establish guidelines for such testing and handle associated risks connected to cybersecurity, radiological, chemical, and biological hazards by the order Biden has signed at the White House. “To realize the promise of AI and avoid the risk, we need to govern this technology,” Biden said. “In the wrong hands, AI can make it easier for hackers to exploit software vulnerabilities that make our society run.”

The Commerce Department said in December 2023 that it was already taking the first steps toward “writing the key standards and guidance for the safe deployment and testing of AI.” The consortium also represents the biggest group of test and evaluation teams  who can now create a foundation for a “new measurement science in AI safety.”

Currently, generative AI has sparked both enthusiasm and concerns with its ability to produce text, images, and videos in response to open-ended cues, that it can eventually replace human labor in some occupations, disrupt elections, and have disastrous consequences.

The Biden administration is working to implement safeguards, but despite multiple high-level conferences, Congress has not passed laws addressing AI.

Featured Image Credit: Photo by Michelangelo Buonarroti; Pexels

The post Address risks: leading AI companies join safety consortium appeared first on ReadWrite.

Iranian state-backed hackers, identified as part of the Islamic Revolutionary Guards, recently disrupted TV streaming services in the United Arab Emirates, according to a recent Guardian report. They broadcasted a deepfake newsreader delivering a fabricated report on the war in Gaza, as reported by Microsoft analysts. This operation, dubbed “For Humanity” by the hackers, involved an AI-generated news anchor presenting unverified images purportedly showing Palestinians harmed by Israeli military actions in Gaza. The Iranian-backed hackers, known as Cotton Sandstorm, showcased their intrusion into three online streaming services on the Telegram messaging platform, interrupting news channels with the fake broadcaster.

In one instance, Dubai residents using a HK1RBOXX set-top box encountered a message claiming the necessity of hacking to deliver a message, followed by the AI-generated anchor introducing “graphic” footage and a ticker detailing casualties in Gaza. The disruptions extended to Canada and the U.K., affecting channels including the BBC, though the BBC itself was not directly hacked.

This incident marks the first time Microsoft has detected an Iranian influence operation leveraging AI as a significant component of its messaging. It represents a notable escalation in the scope of Iranian operations since the onset of the Israel-Hamas conflict, reaching audiences in the UAE, U.K., and Canada.

Deepfakes and election disruption

The rise of generative AI, capable of producing convincing text, voice, and images from simple prompts, has led to an increase in deepfake content online. Such technology poses a risk of being used to disrupt elections, including the upcoming 2024 U.S. presidential election. Iran’s history of targeting the 2020 U.S. election with cyber-campaigns, including impersonating American extremists and spreading disinformation about voting infrastructure, underscores the potential threat posed by these capabilities.

Microsoft’s report highlights the broad range of cyber-attacks and online influence operations launched by Iranian state-backed actors since the Hamas attacks on Oct. 7. These tactics have included exaggerating the impact of cyber-attacks, leaking personal data from an Israeli university, and targeting pro-Israel countries like Albania and Bahrain, as well as the U.S.

The post Iranian hackers broadcast deepfake news in cyber attack on UAE streaming services appeared first on ReadWrite.

Microsoft is ramping up marketing of its artificial intelligence (AI) assistant Copilot with an advertisement designed for the Super Bowl launching on YouTube. While they are not allowed to specifically describe it as a Super Bowl ad, the use of the words “game day” in the title is a big clue.

The ad highlights the multiple ways in which Microsoft expects users to engage with Copilot. It is prominently shown on both a home computer and a mobile phone to illustrate the fact that Copilot is not locked to Microsoft decides and can be used wherever its users are. The closing tagline of the advert is “Anyone. Anywhere. Any device.”

In the advert, users are shown using Copilot for a very wide variety of tasks. Microsoft is trying to create an extremely broad market appeal for the tool. From quizzing someone on their organic chemistry to generating code for a budding video game developer, Copilot seems to have a lot of uses.

Demonstrating to a wide market what an artificial intelligence tool can do for them is a big challenge and if Microsoft’s ‘game day’ ad can show the general population how and why they might want to use an AI, they could see huge uptake on Copilot. The Super Bowl is one of the most-watched televised events so the advert will reach as many as 100 million people or more.

What is Copilot?

Microsoft has been developing its AI assistant for some time, but until recently it was known as Bing Chat. In December, Microsoft integrated OpenAI’s Chat GPT-4 Turbo with its tool and rebranded as Copilot. This brought Copilot much closer to its current feature set by incorporating code interpretation.

Earlier this month, Microsoft launched the Pro version of Copilot. It includes enhanced features and priority access to the newest AI models, ensuring a top-of-the-range experience for paid customers.

On Thursday (Feb.8) Google announced the launch of Gemini Advanced, marking a significant update to its own artificial intelligence technology.

Featured image credit: Microsoft

The post AI has its Super Bowl moment with Microsoft Copilot appeared first on ReadWrite.

Google announced today the launch of Gemini Advanced, marking a significant update to its artificial intelligence technology. This new development is part of Google’s ongoing efforts to integrate AI across its range of products and services, including the Cloud and Workspace services, as well as Google One, which is on the verge of reaching 100 million subscribers.

In December, Google introduced a comprehensive upgrade to its AI models, which have shown remarkable performance across various benchmarks in text, image, audio, and video processing. These models are designed to support not just Google’s consumer products but also to provide developers and businesses with advanced tools for innovation.

Google Bard is now Gemini Advanced

Gemini Advanced, previously known as Bard, represents Google’s latest AI offering, now available in 40 languages and soon to be released on Android and iOS platforms. This enhanced version is capable of complex reasoning, coding, creative tasks, and personalized learning assistance. It’s positioned as a tool for users seeking a more sophisticated AI experience for tasks ranging from content strategy planning to business plan development.

Access to Gemini Advanced is through a subscription to Google One AI Premium plan, which combines Google’s leading AI features into a comprehensive package. This plan extends beyond Google One’s traditional offerings, which include expanded storage and exclusive product features.

Google is also integrating its advanced AI models into everyday products that businesses and individuals frequently use. For Workspace users, over a million people are already utilizing AI to boost productivity and creativity. Google is rebranding the AI features in Workspace, known as Duet AI, as Gemini for Workspace, extending these capabilities to consumers who subscribe to the Google One AI Premium plan.

For Google Cloud customers, the transition to Gemini will occur in the coming weeks, aiming to enhance productivity, accelerate development, and improve cybersecurity for companies.

Google emphasizes the role of developers in the evolution of its AI ecosystem, with a significant number already leveraging Gemini models for their projects. While today’s announcement focused on Gemini Advanced and its capabilities, Google says it plans to reveal more details about offerings for developers and Cloud customers in the near future.

The post Google expands AI offerings with launch of Gemini Advanced appeared first on ReadWrite.

The Apple Vision Pro was launched only a week ago and already we’re beginning to see some exciting apps to go with it. Such as Outtakes, an app announced on Wednesday (Feb 7) that will allow users to create and move around their dream-like worlds.

Utah-based software company, Outtakes broke the news of the launch of its flagship Outtakes App on X. It will run on Mac devices equipped with Apple Silicon chips and creates immersible worlds for wearers of the mixed reality headset.

Breakthrough

Generate entire worlds with Outtakes for  Vision Pro

Immersive animated 360 worlds upscaled to 8K, runs entirely local on Apple Silicon. All you need is a prompt and Outtakes AI will handle the rest.

Coming soon to the App Store. DM for beta. pic.twitter.com/LNjR626ZX5

— Justin Means (@justincmeans) February 7, 2024

Outtakes App uses its generative artificial intelligence (AI) to create “immersive animated 360 worlds” which are then upscaled to 8k. Apple Vision Pro users are immersed in the visual creation through the headset’s display.

Users will need author prompts for Outtakes AI to use as the basis for the generation. It then uses the text=to-image model Stable Diffusion to fulfill the request. This is done on an Apple computer and then sent to the Apple Vision Pro for users to enjoy.

The app is built using Apple’s CoreML, which allows developers to integrate machine learning into their apps. It makes use of the power behind Apple Silicon chips and the Vision Pro’s built-in Neural Engine chip to serve content to users.

CEO of Outtakes Justin Means said in response to questions on Twitter that although the app currently generates flat 360 images using a skybox, both full 3D and volumetric depth maps are on the roadmap for future development of Outtakes.

Means has stated that as it currently is, using Outtakes and an Apple Vision Pro is “a lot like being in the New Vegas Sphere.” One user compared it to Star Trek’s Holodeck, though true interactivity seems to be a long way off. Outtakes App is predominantly visual at this stage of its development.

Many generative AI tools allow users to create images and videos based on prompts, but Outtakes App’s integration with Apple’s native technology and its included editing capabilities are what set it apart from similar competitors.

When is the Outtakes App coming out?

Currently, there is no announced date for the release of Outtakes App, but it is generating some excitement on social media. CEO Means has stated that he is “excited to get this out” and was offering beta access to interested users, implying that the general public should not have to wait too long for the launch.

Featured image credit: Outtakes App screenshot

The post Outtakes AI: App will allow users to generate incredible worlds on Vision Pro. appeared first on ReadWrite.

A group of scientists, doctors, and entrepreneurs have pulled in $12 million in seed investment for a new noninvasive health wearable powered by artificial intelligence (AI).

Created by health tech company Elemind, the wearable neurotechnology platform uses proprietary algorithms and AI to improve health through brainwave stimulation, offering apparent benefits without the side effects of pharmaceuticals. The first round of seed investment has pulled in $12 million for the program, following multiple studies proving its efficacy, it was announced on Tuesday (Feb.6).

How does Elemind work?

The wearable works by reading individual brainwaves and stimulating them in real-time with responses tailored to the individual, thanks to its AI. This ability for precise guidance of brainwaves offers a more targeted approach than traditional pharmaceuticals, as well as fewer side effects.

“Chemical drugs affect the entire body, often leading to unwanted side effects. Elemind offers a non-chemical, direct, and on-demand solution that learns and dynamically adjusts to each person,” said Meredith Perry, CEO and co-founder of Elemind. “We’re the first and only company able to precisely guide and redirect brainwaves in real-time.”

Elemind has dubbed the approach “electric medicine”, created through partnerships with multiple leading research institutions from the Massachusetts Institute of Technology (MIT), Imperial College London, Harvard Medical School and the University of California at Berkeley. Multiple applications have been adapted for the platform, offering various ways to employ it through wearable technology.

Early clinical trials demonstrate that Elemind’s technology is effective at inducing sleep up to 74% faster, decreasing tremors after just 30 seconds of stimulation, and boosting memory. The team behind Elemind hopes for further future “potential impact across many neurological conditions”.

As a general wellness device, Elemind’s first product is not subject to FDA regulation. Thanks to early investment from Village Global, LDV Partners, and other funds, the company plans to build more wearable adaptations for the neurotech platform.

Featured image: Pexels

The post Elemind: Wearable AI offers ‘electric medicine’ via brain stimulation appeared first on ReadWrite.

China is making significant progress in developing advanced semiconductors even as the United States tightens export controls to contain its technological ambitions.

The country’s leading chipmaker SMIC has new production lines in Shanghai to manufacture next-generation 5-nanometer (5nm) mobile processors, reports the Financial Times. The chips are designed by technology giant Huawei as part of its homegrown Kirin line of smartphone processors.

The importance of China’s 5nm chips

The 5nm node represents a significant milestone and lags just one generation behind the cutting-edge 3nm chips from global industry leaders. The process shrinks the smallest features on computer chips to just 5 nanometers wide. It’s exceptionally difficult to do. The technology required to undertake the task successfully is expensive and bleeding edge.

Establishing volume production at this advanced node signals Beijing’s strategic goals of greater self-sufficiency in semiconductors is close to being realized.

The 5nm semiconductor chips are used for high-end smartphone processors and modems as well as artificial intelligence and high-performance computing chips, enabling more powerful and energy-efficient processors for data centers, cloud infrastructure, and specialized AI hardware. They are vital to China’s AI ambitions. 

The collaboration between SMIC and Huawei demonstrates that China’s domestic semiconductor ecosystem retains the ability to innovate at leading-edge process nodes despite U.S. efforts to stymie development through export controls.

Insiders told the Financial Times that if the new 5nm line can successfully manufacture processors for smartphones, SMIC will also produce Huawei’s highest-performance artificial intelligence chip, the Ascend 920. This would help narrow the gap with advanced AI accelerators such as Nvidia’s dominant data center GPUs.

“SMIC is facing a more significant roadblock for production expansion after the US and its alliance tightened export restrictions on advanced chipmaking gear,” said the FT’s source. “Still, the fate of China’s chip industry and its technological development in the coming years will depend on these production lines by SMIC.”

Featured image: Pexels/Ágoston Fung: 

The post China’s SMIC and Huawei building advance 5nm chips despite US appeared first on ReadWrite.