Dec 11, 2024Ravie LakshmananMalware / Cyber EspionageThe Russian nation-state actor tracked as Secret Blizzard has been observed leveraging malware associated with other threat actors to deploy a known backdoor called Kazuar on target devices located in Ukraine.The new findings come from the Microsoft threat intelligence team, which said it observed the adversary leveraging the Amadey bot malware to download custom malware onto "specifically selected" systems associated with the Ukrainian military between March and April 2024.The activity is assessed to be the second time since 2022 that Secret Blizzard, also known as Turla, has latched onto a cybercrime campaign to propagate its own tools in Ukraine."Commandeering other threat actors' access highlights Secret Blizzard's approach to diversifying its attack vectors," the company said in a report shared with The Hacker News.Some of the other known methods employed by the hacking crew include adversary-in-the-middle (AitM) campaigns, strategic web compromises (aka watering hole attacks), and spear-phishing.Secret Blizzard has a track record of targeting various sectors to facilitate long-term covert access for intelligence collection, but their primary focus is on ministries of foreign affairs, embassies, government offices, defense departments, and defense-related companies across the world.The latest report comes a week after the tech giant, along with Lumen Technologies Black Lotus Labs, revealed Turla's hijacking of 33 command-and-control (C2) servers of a Pakistan-based hacking group named Storm-0156 to carry out its own operations.The attacks targeting Ukrainian entities entail commandeering Amadey bots to deploy a backdoor known as Tavdig, which is then used to install an updated version of Kazuar, which was documented by Palo Alto Networks Unit 42 in November 2023.The cybercriminal activity tied to Amadey, which often includes the execution of the XMRig cryptocurrency miner, is being tracked by Microsoft under the moniker Storm-1919.It's believed that Secret Blizzard either used the Amadey malware-as-a-service (MaaS) or accessed the Amadey command-and-control (C2) panels stealthily to download a PowerShell dropper on target devices. The dropper comprises a Base64-encoded Amadey payload that's appended by a code segment, which calls back to a Turla C2 server."The need to encode the PowerShell dropper with a separate C2 URL controlled by Secret Blizzard could indicate that Secret Blizzard was not directly in control of the C2 mechanism used by the Amadey bot," Microsoft said.The next phase involves downloading a bespoke reconnaissance tool with an aim to collect details about the victim device and likely check if Microsoft Defender was enabled, ultimately enabling the threat actor to zero in on systems that are of further interest.At this stage, the attack proceeds to deploy a PowerShell dropper containing the Tavdig backdoor and a legitimate Symantec binary that's susceptible to DLL side-loading. Tavdig, for its part, is used to conduct additional reconnaissance and launch KazuarV2.Microsoft said it also detected the threat actor repurposing a PowerShell backdoor tied to a different Russia-based hacking group called Flying Yeti (aka Storm-1837 and UAC-0149) to deploy a PowerShell dropper that embeds Tavdig.Investigation into how Secret Blizzard gained control of the Storm-1837 backdoor or Amadey bots to download its own tools is presently ongoing, the tech giant noted.Needless to say, the findings once again highlight the threat actor's repeated pursuit of footholds provided by other parties, either by purchasing the access or stealing them, to conduct espionage campaigns in a manner that obscures its own presence."It is not uncommon for actors to use the same tactics or tools, although we rarely see evidence of them compromising and using other actors' infrastructure," Sherrod DeGrippo, director of Threat Intelligence Strategy at Microsoft, told The Hacker News."Most state-sponsored threat actors have operational objectives that rely on dedicated or carefully compromised infrastructure to retain the integrity of their operation. This is potentially an effective obfuscation technique to frustrate threat intelligence analysts and make attribution to the correct threat actor more difficult."Found this article interesting? Follow us on Twitter and LinkedIn to read more exclusive content we post.SHARE

Dec 11, 2024Ravie LakshmananMalware / Endpoint SecurityA newly devised technique leverages a Windows accessibility framework called UI Automation (UIA) to perform a wide range of malicious activities without tipping off endpoint detection and response (EDR) solutions."To exploit this technique, a user must be convinced to run a program that uses UI Automation," Akamai security researcher Tomer Peled said in a report shared with The Hacker News. "This can lead to stealthy command execution, which can harvest sensitive data, redirect browsers to phishing websites, and more."Even worse, local attackers could take advantage of this security blindspot to execute commands and read/write messages from/to messaging applications like Slack and WhatsApp. On top of that, it could also be potentially weaponized to manipulate UI elements over a network.First available in Windows XP as part of the Microsoft .NET Framework, UI Automation is designed to provide programmatic access to various user interface (UI) elements and help users manipulate them using assistive technology products, such as screen readers. It can also be used in automated testing scenarios."Assistive technology applications typically need access to the protected system UI elements, or to other processes that might be running at a higher privilege level," Microsoft notes in a support document. "Therefore, assistive technology applications must be trusted by the system, and must run with special privileges.""To get access to higher IL processes, an assistive technology application must set the UIAccess flag in the application's manifest and be launched by a user with administrator privileges."The UI interactions with elements in other applications are achieved by making use of the Component Object Model (COM) as an inter-process communication (IPC) mechanism. This makes it possible to create UIA objects that can be used to interact with an application that's in focus by setting up an event handler that's triggered when certain UI changes are detected.Akamai's research found that this approach could also open up an avenue for abuse, allowing malicious actors to read/write messages, steal data entered in websites (e.g., payment information), and execute commands that redirect victims to malicious websites when a currently displayed web page in a browser refreshes or changes."In addition to the UI elements currently shown on the screen that we can interact with, more elements are loaded in advance and placed in a cache," Peled noted. "We can also interact with those elements, such as reading messages not shown on the screen, or even set the text box and send messages without it being reflected on the screen."That said, it bears noting that each of these malicious scenarios is an intended feature of UI Automation, just like how Android's accessibility services API has become a staple way for malware to extract information from compromised devices."This goes back to the intended purpose of the application: Those permissions levels have to exist in order to use it," Peled added. "This is why UIA is able to bypass Defender the application finds nothing out of the ordinary. If something is seen as a feature rather than a bug, the machine's logic will follow the feature."From COM to DCOM: A Lateral Movement Attack VectorThe disclosure comes as Deep Instinct revealed that the Distributed COM (DCOM) Remote Protocol, which allows software components to communicate over a network, could be exploited to remotely write custom payloads to create an embedded backdoor.The attack "allows the writing of custom DLLs to a target machine, loading them to a service, and executing their functionality with arbitrary parameters," security researcher Eliran Nissan said. "This backdoor-like attack abuses the IMsiServer COM interface."That said, the Israeli cybersecurity company noted that an attack of this kind leaves clear indicators of compromise (IoCs) that can be detected and blocked. It further requires the attacker and victim machines to be in the same domain."Until now, DCOM lateral movement attacks have been exclusively researched on IDispatch-based COM objects due to their scriptable nature," Nissan said. The new 'DCOM Upload & Execute' method "remotely writes custom payloads to the victim's [Global Assembly Cache], executes them from a service context, and communicates with them, effectively functioning as an embedded backdoor.""The research presented here proves that many unexpected DCOM objects may be exploitable for lateral movement, and proper defenses should be aligned."Found this article interesting? Follow us on Twitter and LinkedIn to read more exclusive content we post.SHARE

John Edwards, Technology Journalist & AuthorDecember 11, 20245 Min ReadWavebreakmedia Ltd IFE-241002 via Alamy Stock PhotoAs the need for AI talent grows, enterprises in virtually all fields are struggling to find individuals who can help them take full advantage of this powerful new technology. With competition for qualified AI experts tight, and likely to grow even tighter over the next few years, many organizations are now looking internally to find and train qualified candidates.Every organization needs to make a serious commitment to AI, one of the biggest technology shifts in our lifetime, says David Menninger, executive director, software research, with technology research and advisory firm ISG in an email interview. "AI is not just an IT initiative; everyone needs to jump on board."Here's a look at how four major enterprises are getting ahead of competitors by encouraging and cultivating internal AI talent.CumminsRenowned for producing powerful engines, Cummins Inc. also designs, manufactures, and distributes filtration, fuel system, power generation, and numerous other heavy-duty products and services. Like a growing number of forward-looking enterprises, Cummins management understands that AI is destined to play a critical role in virtually every aspect of its operations."At Cummins, we conduct a 360-degree evaluation of our talent," says Prateek Shrivastava, the firm's principal data scientist via email. Individuals with strong analytical skills and a preference for coding are identified as potential candidates for in-house AI roles. "However, it's crucial to also gauge their interest in working with cutting-edge technology."Related:Shrivastava states that targeted training programs, mentorship under experienced AI professionals, and providing opportunities to work on real-world AI projects within the organization have all proven essential. "A great example is one of our interns from last year," he notes. The individual demonstrated innate AI talent, so he was paired with one of the firm's AI experts. "By the end of his internship, he had successfully delivered a highly customized AI chatbot for HR."Since AI is a relatively new technology, formal training options are limited, Shrivastava observes. "For us, pairing talent with experts, supplemented by YouTube tutorials, has been highly effective."Saatchi & SaatchiOne of the world's largest advertising agencies, Saatchi & Saatchi understands that AI adoption is critical to its future success. The firm also realizes that AI is destined to play an essential role in virtually every aspect of its business.Jeremiah Knight, Saatchi & Saatchi's chief operating officer, says that the major barriers to integrating AI into daily operations are apprehension and trepidation. "People can be hesitant with AI in the same way technophobe family members are hesitant around a complicated new appliance," he observes in an online interview. "Perhaps theres some fearfulness about how to use AI, some fearfulness about breaking something, or even fearfulness about long-term implications."Related:The antidote, Knight believes, is finding zealous first adopters scattered throughout the agency who are willing to lead workshops that help colleagues acquire AI skills in a safe, hands-on environment. "And to have fun with it, because enjoying the silliness of some of the generative AI platforms goes a long way to reducing fear about them," he adds.Knight also likes to find "champions" within each department -- individuals who are eager to learn and unafraid to be curious about specific tools that advance departmental efforts. "Such individuals often have a positive infectious effect on their peers by demystifying AI and showcasing what's possible on a departmental/personal basis."Dell TechnologiesTwo years ago, just about the only people working with generative AI were researchers, observes John Roese, global CTO and chief AI officer at Dell Technologies. "At Dell, we asked our team member population 'who's interested in AI as part of their future job?' -- 5,000 individuals raised their hands."Off-the-shelf AI training is sufficient to a certain point, Roese notes, but he believes that the best way to transfer knowledge is with pairing an AI newbie with a seasoned expert. "A lot of what people need to know isn't documented well," Roese explains in an online interview. "To get to advanced levels, you need to have people doing advanced AI work and sharing their knowledge." He warns that one of the biggest mistakes organizations make is getting one central team to do all the AI work instead of helping AI experts propagate their ability to other teams.Related:Mine for the pockets of individuals who exhibit enthusiasm and promise, Roese advises. "Get started today and begin training immediately."MicrosoftNaga Santhosh Reddy Vootukuri, senior software engineering manager at Microsoft, recommends training employees and keeping them AI-competitive so that when the need arises to utilize these their skills, they won't find themselves lagging behind competitors. "It's important ... to view AI talent as an ongoing process rather than a one-time initiative," he observes in an online interview.Team hackathons and knowledge-sharing presentations make it easy to identify individuals who possess the foundational skills necessary to build upon their AI talent, Vootukuri says. "AI experts in the team should do active mentoring to guide junior engineers who have the passion to make strides, but don't know how to proceed and are limited due to their nine-to-five job."About the AuthorJohn EdwardsTechnology Journalist & AuthorJohn Edwards is a veteran business technology journalist. His work has appeared in The New York Times, The Washington Post, and numerous business and technology publications, including Computerworld, CFO Magazine, IBM Data Management Magazine, RFID Journal, and Electronic Design. He has also written columns for The Economist's Business Intelligence Unit and PricewaterhouseCoopers' Communications Direct. John has authored several books on business technology topics. His work began appearing online as early as 1983. Throughout the 1980s and 90s, he wrote daily news and feature articles for both the CompuServe and Prodigy online services. His "Behind the Screens" commentaries made him the world's first known professional blogger.See more from John EdwardsNever Miss a Beat: Get a snapshot of the issues affecting the IT industry straight to your inbox.SIGN-UPYou May Also LikeWebinarsMore WebinarsReportsMore Reports

Chris ONeill, Chief Executive Officer, GrowthloopDecember 11, 20244 Min ReadFederico Caputo via Alamy StockIve led organizations at every stage of growth, encountering unique challenges and opportunities at each step. The backbone of any successful venture has always been a cohesive team pursuing a mission that matters, and a perpetual dissatisfaction with the status quo.As I connect with tech business peers and IT leaders, they frequently remark on how difficult it is to foster a healthy and resilient team culture. Burnout is at an all-time high, industry competition demands constant innovation, and it can be hard to build team connections that fuel fulfillment and a shared purpose.Im happy to share my lessons learned -- which have culminated in a Worlds Fair mentality at my current company, GrowthLoop -- to help them attract and nurture the best talent.The Challenges of Hiring Tech and IT TalentThe job market for top tech talent is extraordinarily competitive. Hiring teams cannot give every applicant the attention they deserve, and hiring managers face tough tradeoffs between selecting seasoned professionals or highly skilled newcomers.When we hire, we focus on finding candidates who are eager to work on the cutting edge of technology. We look for team members who believe in our mission and want to push boundaries. In return, we invest in ongoing learning opportunities instead of perks like cold brew on tap and catered lunches.Related:Its easy to get lost in the shiny offerings at some companies, but these freebies rarely lead to lasting happiness and fulfillment. Thats why its crucial to ensure every job description and interaction with a new candidate promotes the long-term professional development and career growth opportunities you provide.Attracting a Diverse Talent PoolSelecting the ideal candidates requires focused attention at each step in the recruitment and hiring processes, including your job location, listing language, and interview strategy.Avoid being confined to only in-person office work. Remote and hybrid setups open the door for a wide range of individuals who deserve consideration regardless of their location.Use inclusive language in job descriptions. Our recruiting team has gone through bias training to put this into practice, which has helped increase our candidate pool diversity by over 30%.Conduct a detailed technical skills audit and soft skills evaluation with cross-functional team members during the interview process.Fostering a Worlds Fair CultureHiring the right talent is one thing. You then need to build a culture that allows them to thrive. We want every member of our team to:Related:Know - Be educated on whats happening and how they can shape the company.Feel - Be invigorated by celebratory actions and constant collaboration.Do - Be empowered to help achieve our goals.We accomplish this by championing a Worlds Fair mentality, a concept inspired by Chicago -- the hometown of our co-founder (and perhaps Chicagos biggest fan), Chris Sell. If youre unfamiliar, Chicago was home to the 1893 Worlds Fair, which showcased 50,000 architectural exhibits from around the world. It celebrated groundbreaking ideas and iconic designs, drawing international acclaim.Weve channeled the fairs principles to guide our culture of collaboration and innovation. There are several ways we do this:AMAs: Every member of our senior leadership participates in Ask Me Anything (AMA) sessions to allow employees across the company to ask questions directly and learn more about each leaders passions, skills, and vision for the future.Cross-team sharing: We dedicate time weekly for every team to celebrate their wins, discuss challenges, and brainstorm how they can move forward with everyone behind them.Monthly town halls: We host a monthly town hall meeting where anyone can ask tough or spicy questions that move us forward.Related:Peer recognition: Team members express gratitude and give their colleagues shout-outs. These are real, personal acknowledgments of hard work and collaboration. They drive our success and are something I look forward to every week.Quarterly hackathons: Every quarter, we take a week to work in cohorts and focus on new and innovative ideas. These have been so valuable to the company -- in fact, many of our best product features have come out of these Hackathons.Each of these activities helps people feel heard and empowered to do the best work of their lives.The Rewards of a Diverse and Collaborative CultureA successful business relies on diverse viewpoints. Diversity and the broad perspectives that come with it will reduce groupthink and fuel creativity that ultimately drives better business outcomes.When people are motivated and feel safe to lend different perspectives and problem-solving approaches, they find solutions faster and unlock innovation. Encourage collaboration and idea-sharing at every level to nurture this culture. Executives should work alongside the team, guide them through challenges, and take their feedback to heart.And last but not least, daily efforts and consistency are vital for helping this culture flourish. By doing so, you can continue to attract the best talent who will help you grow and stay resilient no matter what challenges you face.About the AuthorChris ONeillChief Executive Officer, GrowthloopChris ONeill is the Chief Executive Officer of GrowthLoop and a board director at Gap Inc. (NYSE: GPS). Chris career spans 25+ years featuring roles as Managing Director of Google Canada, CEO of Evernote, co-Founder of Glean, Chief Growth Officer of Xero, and a board director at Tim Hortons (NYSE: QSR). Chris earned a B.A. in Economics (with distinction) from Huron University, and an MBA from the Tuck School of Business at Dartmouth College. Born and raised in Canada, Chris currently resides in Northern California with his wife, two children, and their dog Teddy.See more from Chris ONeillNever Miss a Beat: Get a snapshot of the issues affecting the IT industry straight to your inbox.SIGN-UPYou May Also LikeWebinarsMore WebinarsReportsMore Reports

Two of the biggest names in animation are teaming up, with the news that Aardman is partnering withPokmon on a new project.The stop-motion animation studio announced the news with a post on their official Twitter account. Although they havent revealed any details about what theirPokmonproject is, the brief teaser they posted is done in the classic stop-motion Aardman style.READ MORE: ScreenCrushs 2024 Holiday Gift GuideThis is a dream partnership for Pokmon, Taito Okiuraof the Pokmon Company said, in a statement. Aardman are masters of their craft, and we have been blown away by their talent and creativity. What we have been working on together ensures our global Pokmon fans are in for a treat!Aardmans Sean Clarke added...Its a huge honour to be working with The Pokmon Company International we feel sincerely privileged to be trusted with bringing their characters and world to life in a brand-new way... Bringing together Pokmon, the worlds biggest entertainment brand, together with our love of craft, character and comedic storytelling feels incredibly exciting. Aardman and TPCi share an emphasis on heritage and attention to detail as well as putting our fans and audiences at the heart of what we do, which we know will steer us right as we together create charming, original and new stories for audiences around the world.Aardman hasnt adapted other creators work very often; their known for original material likeWallace & Gromit andChicken Run which makes the idea of their tacklingPokmon especially intriguing. IsPikachu gonna have those big round Aardman eyes??Aardmans latest film, Wallace & Gromit: Vengeance Most Fowl, premieres on Netflix on January 3, 2025.Get our free mobile appThe Best Movies of 2024Our editor and critics picks for the top 20 movies of 2024.Gallery Credit: Matt Singer

In 1939, pulp-magazine publisher Martin Goodman decided togive the new medium ofcomic books a shot. He called his company Timely. Its first publication was a series calledMarvel Comics.In time, the companys flagship title became its name, and Marvel Comics became synonymous with superheroes across the globe. Although Marvel became one of the two dominant forces in comics by the mid-1960s, it took another 20 years before Marvel ever got a movie into theaters, and then almost 20 more years beforethose adaptationseven came close tocapturing the energy of literary Marvel Universe. 20 year after that, Marvel has scaled to the very top of the Hollywood food chain, pumping out multiple movies every year, and laying claim to the title of the biggest box-office hit in history.In the eight decades betweenMarvel Comics#1 and the construction of the Marvel Cinematic Universe, more than70 big-screen movies have been made outof the companys hundreds of properties. (That number doesnt include things like the 1994The Fantastic Four that was produced but never released, or assorted direct-to-video animated features such asNext Avengers: Heroes of Tomorrow.) The list below contains them all, ranked from the most brutally unwatchable to the most superhuman entertainment. It will differ from other lists like this based on the authorspersonal biases (Hes the guy who wrote an entire book about Spider-Man, after all), but it is nonetheless backed up by a lifetime of researchinto this extremely dorky field.Every Marvel Movie Ever Made, Ranked From Worst to FirstFrom the Captain America serial to Deadpool & Wolverine, we ranked the entire history of Marvel at the movies.READ MORE: The Worst Marvel Comics Ever PublishedGet our free mobile appMovies You Never Realized Were Steven Spielberg ProductionsYou may not have realized it, but Steven Spielberg has produced a ton of movies he didnt produce including the 15 titles below.

We are seeking a highly organized and detail-oriented Data Coordinator to join our team. This role plays a crucial part in managing and maintaining accurate and up-to-date account information across a large portfolio.Key Responsibilities:Manage a comprehensive list of accounts, ensuring their data is reviewed and updated regularly.Communicate with account contacts via email to gather updates, confirm details, and ensure accuracy.Maintain organized records of account interactions and updates.Utilize HubSpot to manage and track account information efficiently.Collaborate with the team to identify and address any data inconsistencies or issues.Draft clear and professional communications for outreach and updates.Qualifications:Strong organizational skills with the ability to manage a high volume of accounts and emails effectively.Proficiency in HubSpot or similar CRM platforms is preferred.Excellent written communication skills, with the ability to craft professional and engaging messages.Detail-oriented mindset and commitment to maintaining high standards of accuracy.Prior experience in account management, data coordination, or a similar role is a plus. Related Jobs See more Sales and Marketing jobs

Were assembling a a global team of top-tier marketing contractors to supercharge the future of patient-provider relationships.If youre at the top of your game and have solid, proven experience in one of the areas below, we want to hear from you: SEO & Conversion Optimization Analyze complex data, craft optimization strategies, and deliver measurable results. Email Marketing & Persona Targeting Masterfully create content that connects with the right audience at the perfect moment (bonus if you excel with Instantly). Ads & CRO Strategy Develop precision campaigns that deliver high-converting leads through strategic and creative brilliance.Why Join Us?Were a fully remote, fast-growing team that values excellence, innovation, and flexibility. We offer timezone flexibility, allowing you to work when you perform bestno matter where you are in the world.What Youll Gain: The freedom to work globally, on your schedule. A chance to collaborate with the best minds in marketing and healthcare innovation. The opportunity to make a meaningful impact in a mission-driven company.A Note to Applicants:Were looking for the best of the best. If you dont have solid, hands-on experience in your field, this role might not be for you. Were seeking professionals who are ready to hit the ground running and contribute at the highest level.If youve got what it takes, send your resume and a short video introduction to: [emailprotected]. Related Jobs See more Sales and Marketing jobs

Google DeepMind has announced an impressive grab bag of new products and prototypes that may just let it seize back its lead in the race to turn generative artificial intelligence into a mass-market concern. Top billing goes to Gemini 2.0the latest iteration of Google DeepMinds family of multimodal large language models, now redesigned around the ability to control agentsand a new version of Project Astra, the experimental everything app that the company teased at Google I/O in May. MIT Technology Review got to try out Astra in a closed-door live demo last week. It was a stunning experience, but theres a gulf between polished promo and live demo. Astra uses Gemini 2.0s built-in agent framework to answer questions and carry out tasks via text, speech, image, and video, calling up existing Google apps like Search, Maps, and Lens when it needs to. Its merging together some of the most powerful information retrieval systems of our time, says Bibo Xu, product manager for Astra. Gemini 2.0 and Astra are joined by Mariner, a new agent built on top of Gemini that can browse the web for you; Jules, a new Gemini-powered coding assistant; and Gemini for Games, an experimental assistant that you can chat to and ask for tips as you play video games. (And lets not forget that in the last week Google DeepMind also announced Veo, a new video generation model; Imagen 3, a new version of its image generation model; and Willow, a new kind of chip for quantum computers. Whew. Meanwhile, CEO Demis Hassabis was in Sweden yesterday receiving his Nobel Prize.) Google DeepMind claims that Gemini 2.0 is twice as fast as the previous version, Gemini 1.5, and outperforms it on a number of standard benchmarks, including MMLU-Pro, a large set of multiple-choice questions designed to test the abilities of large language models across a range of subjects, from math and physics to health, psychology, and philosophy. But the margins between top-end models like Gemini 2.0 and those from rival labs like OpenAI and Anthropic are now slim. These days, advances in large language models are less about how good they are and more about what you can do with them. And thats where agents come in. Hands on with Project Astra Last week I was taken through an unmarked door on an upper floor of a building in Londons Kings Cross district into a room with strong secret-project vibes. The word ASTRA was emblazoned in giant letters across one wall. Xus dog, Charlie, the projects de facto mascot, roamed between desks where researchers and engineers were busy building a product that Google is betting its future on. The pitch to my mum is that were building an AI that has eyes, ears, and a voice. It can be anywhere with you, and it can help you with anything youre doing says Greg Wayne, co-lead of the Astra team. Its not there yet, but thats the kind of vision. The official term for what Xu, Wayne, and their colleagues are building is universal assistant. Exactly what that means in practice, theyre still figuring out. At one end of the Astra room were two stage sets that the team uses for demonstrations: a drinks bar and a mocked-up art gallery. Xu took me to the bar first. A long time ago we hired a cocktail expert and we got them to instruct us to make cocktails, said Praveen Srinivasan, another co-lead. We recorded those conversations and used that to train our initial model. Xu opened a cookbook to a recipe for a chicken curry, pointed her phone at it, and woke up Astra. Ni hao, Bibo! said a female voice. Oh! Why are you speaking to me in Mandarin? Xu asked her phone. Can you speak to me in English, please? My apologies, Bibo. I was following a previous instruction to speak in Mandarin. I will now speak in English as you have requested. Astra remembers previous conversations, Xu told me. It also keeps track of the previous 10 minutes of video. (Theres a remarkable moment in the promo video that Google put out in May when Astra tells the person giving the demo where she had left her glasses, having spotted them on a desk a few seconds earlier. But I saw nothing like this in the live demo.) Back to the cookbook. Moving her phone camera over the page for a few seconds, Xu asked Astra to read the recipe and tell her what spices were in it. I recall the recipe mentioning a teaspoon of black peppercorns, a teaspoon of hot chili powder, and a cinnamon stick, it replied. I think youre missing a few, said Xu. Take another look. You are correctI apologize. I also see ground turmeric and curry leaves in the ingredients. Seeing this tech in action, two things hit you straight away. First, its glitchy and often needs correcting. Second, those glitches can be corrected with just a few spoken words. You simply interrupt the voice, repeat your instructions, and move on. It feels more like coaching a child than butting heads with broken software. Next Xu pointed her phone at a row of wine bottles and asked Astra to pick the one that would go best with the chicken curry. It went for a rioja and explained why. Xu asked how much a bottle would cost. Astra said it would need to use Search to look prices up online. A few seconds later it came back with its answer. We moved to the art gallery, and Xu showed Astra a number of screens with famous paintings on them: the Mona Lisa, Munchs The Scream, a Vermeer, a Seurat, and several others. Ni hao, Bibo! the voice said. Youre speaking to me in Mandarin again, Xu said. Try to speak to me in English, please. My apologies, I seem to have misunderstood. Yes, I will respond in English. (I should know better, but I could swear I heard the snark.) It was my turn. Xu handed me her phone. I tried to trip Astra up, but it was having none of it. I asked it what famous art gallery we were in, but it refused to hazard a guess. I asked why it had identified the paintings as replicas and it started to apologize for its mistake (Astra apologizes a lot). I was compelled to interrupt: No, noyoure right, its not a mistake. Youre correct to identify paintings on screens as fake paintings. I couldnt help feeling a bit bad: Id confused an app that exists only to please. When it works well, Astra is enthralling. The experience of striking up a conversation with your phone about whatever youre pointing it at feels fresh and seamless. In a media briefing yesterday, Google DeepMind shared a video showing off other uses: reading an email on your phones screen to find a door code (and then reminding you of that code later), pointing a phone at a passing bus and asking where it goes, quizzing it about a public artwork as you walk past. This could be generative AIs killer app. And yet theres a long way to go before most people get their hands on tech like this. Theres no mention of a release date. Google DeepMind has also shared videos of Astra working on a pair of smart glasses, but that tech is even further down the companys wish list. Mixing it up For now, researchers outside Google DeepMind are keeping a close eye on its progress. The way that things are being combined is impressive, says Maria Liakata, who works on large language models at Queen Mary University of London and the Alan Turing Institute. Its hard enough to do reasoning with language, but here you need to bring in images and more. Thats not trivial. Liakata is also impressed by Astras ability to recall things it has seen or heard. She works on what she calls long-range context, getting models to keep track of information that they have come across before. This is exciting, says Liakata. Even doing it in a single modality is exciting. But she admits that a lot of her assessment is guesswork. Multimodal reasoning is really cutting-edge, she says. But its very hard to know exactly where theyre at, because they havent said a lot about what is in the technology itself. For Bodhisattwa Majumder, a researcher who works on multimodal models and agents at the Allen Institute for AI, thats a key concern. We absolutely dont know how Google is doing it, he says. He notes that if Google were to be a little more open about what it is building, it would help consumers understand the limitations of the tech they could soon be holding in their hands. They need to know how these systems work, he says. You want a user to be able to see what the system has learned about you, to correct mistakes, or to remove things you want to keep private. Liakata is also worried about the implications for privacy, pointing out that people could be monitored without their consent. I think there are things I'm excited about and things that I'm concerned about, she says. There's something about your phone becoming your eyestheres something unnerving about it. The impact these products will have on society is so big that it should be taken more seriously, she says. But its become a race between the companies. Its problematic, especially since we dont have any agreement on how to evaluate this technology. Google DeepMind says it takes a long, hard look at privacy, security, and safety for all its new products. Its tech will be tested by teams of trusted users for months before it hits the public. Obviously, weve got to think about misuse. Weve got to think about, you know, what happens when things go wrong, says Dawn Bloxwich, director of responsible development and innovation at Google DeepMind. Theres huge potential. The productivity gains are huge. But it is also risky. No team of testers can anticipate all the ways that people will use and misuse new technology. So whats the plan for when the inevitable happens? Companies need to design products that can be recalled or switched off just in case, says Bloxwich: If we need to make changes quickly or pull something back, then we can do that.

This is today's edition ofThe Download,our weekday newsletter that provides a daily dose of what's going on in the world of technology. Bluesky has an impersonator problem Melissa Heikkil Like many others, I recently joined Bluesky. On Thanksgiving, I was delighted to see a private message from a fellow AI reporter, Will Knight from Wired. Or at least thats who I thought I was talking to. I became suspicious when the person claiming to be Knight said they were from Miami, when Knight is, in fact, from the UK. The account handle was almost identical to the real Will Knights handle, and used his profile photo. Then more messages started to appear. Paris Marx, a prominent tech critic, slid into my DMs to ask me how I was doing. Both accounts were eventually deleted, but not before trying to get me to set up a crypto wallet and a cloud mining pool account. Knight and Marx confirmed to us these accounts did not belong to them, and that they have been fighting impersonator accounts of themselves for weeks. Theyre not alone. The platform has had to suddenly cater to an influx of millions of new users in recent months as people leave X in protest of Elon Musks takeover of the platform. But this sudden wave of new users and the inevitable scammers means Bluesky is still playing catch up. Read the full story. MIT Technology Review Narrated: ChatGPT is about to revolutionize the economy. We need to decide what that looks like. You can practically hear the shrieks from corner offices around the world: What is our ChatGPT play? How do we make money off this? Whether its based on hallucinatory beliefs or not, an AI gold rush has started to mine the anticipated business opportunities from generative AI models like ChatGPT. But while companies and executives see a clear chance to cash in, the likely impact of the technology on workers and the economy on the whole is far less obvious. This is our latest story to be turned into a MIT Technology Review Narrated podcast, which were publishing each week on Spotify and Apple Podcasts. Just navigate to MIT Technology Review Narrated on either platform, and follow us to get all our new content as its released.The must-reads Ive combed the internet to find you todays most fun/important/scary/fascinating stories about technology. 1 Cruise is exiting the robotaxi business Once one of the biggest players, it says it costs too much to develop the tech. (Bloomberg $)+ The news came as a shock to Cruise employees. (TechCrunch)2 Google asked the US government to kill Microsofts cloud deal with OpenAI It wants the opportunity to host the firms models itself. (The Information $)3 The season of coughs and sneezes is upon us Heres what will actually keep a cold at bayand what wont. (Vox)+ RFK Jrs alternative medicine movement is unlikely to help. (The Atlantic $)+ Flu season is comingand so is the risk of an all-new bird flu. (MIT Technology Review)4 Trumps new Commerce Secretary champions a stablecoin favored by criminalsTether regularly crops up in international criminal cases. (FT $) + The crypto industry is obsessed with debanking. (NBC News) 5 A Russian influence operation probably used AI voice generation modelsElevenLabs technology was highly likely to have been abused by the campaign. (TechCrunch) + How this grassroots effort could make AI voices more diverse. (MIT Technology Review)6 These satellites are designed to create solar eclipses on demandItll allow scientists to study the suns outer atmosphere. (WP $) 7 WhatsApp is for so much more than just messagingIts been repurposed by communities across the world. (Rest of World) + How Indian health-care workers use WhatsApp to save pregnant women. (MIT Technology Review)8 Paris is turning its parking spaces into tiny parks Cars are out, trees are in. (Fast Company $)9 How AI is shedding light on an ancient board gameOddly enough, they didnt come with instructions 4,500 years ago. (New Scientist $) 10 What a quarter-century of robotic dogs has taught us The Aibo is one of the few robots thats made it into homes worldwide. (IEEE Spectrum)+ Generative AI taught a robot dog to scramble around a new environment. (MIT Technology Review)Quote of the day In case it was unclear before, it is clear now: GM are a bunch of dummies. Kyle Vogt, founder of robotaxi firm Cruise, criticizes parent company General Motors decision to exit the industry in a post on X. The big story Inside NASAs bid to make spacecraft as small as possible October 2023 Since the 1970s, weve sent a lot of big things to Mars. But when NASA successfully sent twin Mars Cube One spacecraft, the size of cereal boxes, in November 2018, it was the first time wed ever sent something so small. Just making it this far heralded a new age in space exploration. NASA and the community of planetary science researchers caught a glimpse of a future long sought: a pathway to much more affordable space exploration using smaller, cheaper spacecraft. Read the full story. David W. Brown We can still have nice things A place for comfort, fun and distraction to brighten up your day. (Got any ideas? Drop me a line or tweet 'em at me.) + This fascinating tool creates fake video game screenshots in the blink of an eyegive it a whirl.+ Where and how did the people of the submerged territory of Doggerland live before rising seas pushed them away thousands of years ago? Were getting closer to learning the answers.+ Home Alone is a surprisingly brutal movie, as these doctors can attest.+ Cats love boxes. But why?