Latest news with #Veo2
WIRED
a day ago
- Business
- WIRED
I Converted My Photos Into Short Videos With AI on Honor's Latest Phones. It's Weird
The photo gallery app in Honor's latest midrange phones has an image-to-video generative AI feature powered by Google. It'll probably come to your phone soon. All products featured on WIRED are independently selected by our editors. However, we may receive compensation from retailers and/or from purchases of products through these links. As midrange phones designed to plug the gap between flagships, the Honor 400 and 400 Pro might not ordinarily attract much attention. But these devices—unavailable in the US—are among the first to feature Google's image-to-video AI generator, based on its Veo 2 model (now available to Gemini subscribers). Built into Honor's Gallery app, you can select a still photo from your camera roll to bring it to life as a five-second video. After much experimentation with different photos, from landscapes to family and pets, I'm impressed and weirded out. Like any AI tool, it has the potential to be good or bad, depending on how you wield it, and the results veer from flawless to freaky. It's a neat trick, and it's coming to the phone in your hand soon. Fake Photography Faking photography is nothing new—the medium is always evolving. Artificial intelligence has been smoothing wrinkles and enhancing skies for years. None of your photos are real, especially those you shot with your phone and edited to post on social media. But we're pushing beyond creating a fake bokeh background blur or dialing up colors of the sunset. Creating entirely fake videos from still images feels like a new high and a new low. The process is easy. Open the Gallery app on the Honor 400 or 400 Pro, choose the Create tab, tap Image to video, and select one of your photos. Choose a 9:16 or 16:9 aspect ratio for portrait or landscape, then hit start. You need to be connected to the internet. Each five-second video took around 30 seconds to create, but a pop-up message warns me they can take up to two minutes. There's no room to enter a prompt, so you are left at the mercy of whatever the AI decides to do. I began with photos of my wife and kids. The first few videos have a major uncanny valley feel. In one photo, my wife is covering her mouth, and the AI animates her moving her hand and talking, but the mouth it pastes in is entirely wrong. Much to her horror, it gives my daughter a series of facial tics. The video of my selfie comes out well and would surely fool anyone who doesn't know me, but my wife says she can tell it's not me because I never make facial expressions like that. We tried the cats next, and despite some odd expressions (my eldest cat, Bodhi, never looks as deferential as this), the results were pretty solid. Photos of landscapes, such as a boat on the waves or trees in a forest, are brought to life with believable rippling effects. Things got weirder with inanimate objects. With a pair of Star Wars Funko Pops, Anakin Skywalker's lightsaber glowed to life, but Darth Maul on a speeder weirdly rotated on the spot. For a few photos of ornaments, it did a disappointing pan or zoom. As the debate about AI-generated art rages on, I can't help but wonder at the implications of uploading personal photos to be analyzed and animated by ever more powerful AI. Social media is already awash with carefully edited photos and videos, but AI processing is improving fast, generating more believable results with every passing day, and it's quickly becoming accessible to everyone. Camera Cold War Photography is the battleground for phone manufacturers trying to outdo one another. While most phone makers focused on hardware, Google dived deeper into computational photography, doing more with less. Google pioneered and popularized AI features in its Pixel phones. Options like AI Eraser to remove unwanted items or people from photos and AI-enhanced zoom to create more detailed zoomed shots from afar with underpowered hardware are fast becoming ubiquitous, and both work well on Honor's latest phones. Google has already moved on with Pixel 9 features like Best Take, enabling you to pick smiling faces where everyone has their eyes open from different shots, and Add Me, allowing you to take group photos and then add yourself to them. Securing buy-in for this brave new world of AI photography is clearly important, as Google aims to pull big Android phone manufacturers aboard, carefully balancing exclusive Pixel features with debuts for companies like Honor. Honor is creeping closer to marrying hardware and software in the 400 Pro. It has a triple-lens main camera comprising a 200-megapixel main shooter with a large 1/1.4-inch sensor, a 50-megapixel telephoto lens, and a 12-megapixel ultrawide, not to mention a 50-MP selfie camera. Honor has also bet big on AI in recent years, pushing more AI into its phones, working closely with Google, and showing off its own new AI features. The Chinese phone maker has come a long way from beauty filters that smooth out your wrinkles, though those are still available. The generate button has a label that says 'Limited-time free trial.' Honor 400 buyers get two free months and can generate up to 10 videos a day. Presumably, this will eventually be a paid feature of Google's Gemini Pro subscription and will roll out on many more phones (there's no firm timeline or pricing yet). It's a win-win for Google when it owns the underlying AI, enticing you to subscribe down the line. It's fun to play with—we exhausted our daily allowance fast—but seriously creepy.
India Today
5 days ago
- Business
- India Today
Google Veo 3 now available in 71 more countries, Gemini Pro users get trial pack
Google's AI video generator, Veo 3, is being rolled out to users in 71 more countries just days after its initial debut. The announcement came from Josh Woodward, Vice President of Gemini at Google, via a post on X (formerly Twitter). While many regions have now gained access to the tool, countries in the European Union remain notably absent from this rollout. When asked by an X user about India not being on the initial list of countries to get Veo 3, Woodward said, 'Working to enable India as fast as we can!'. For those who don't know, Veo 3 can not only generate visuals from text but also synchronise them with lifelike audio. Early adopters have already begun sharing stunning demo clips online, showing how Veo 3 responds with surprising accuracy to both creative and complex prompts. Source: Josh Woodward/ X (formerly Twitter) advertisementUsers subscribed to the Gemini Pro plan will receive a one-time trial pack that includes ten video generations via the web version of Gemini. Those on the higher-tier Ultra subscription — priced at $249.99 (roughly Rs 21,200) per month — gain full access, with daily generation limits and a dedicated Flow mode. Flow is aimed at video creators and comes with 125 generations monthly for Ultra users, while Pro subscribers stick to their ten. Though the rollout is wide-reaching, Veo 3 remains web-only for now and supports English audio output exclusively. When users upload their own images in Flow mode, voice output is not currently the 71 new regions now included are American Samoa, Angola, Antigua and Barbuda, Argentina, Australia, the Bahamas, Belize, Benin, Bolivia, Botswana, Brazil, Burkina Faso, Cabo Verde, Cambodia, Cameroon, Canada, Chile, Cte d'Ivoire, Colombia, Costa Rica, the Dominican Republic, Ecuador, El Salvador, Fiji, Gabon, Ghana, Guam, Guatemala, Honduras, Jamaica, Japan, Kenya, Laos, Malaysia, Mali, Mauritius, Mexico, Mozambique, Namibia, Nepal, New Zealand, Nicaragua, Niger, Nigeria, the Northern Mariana Islands, Pakistan, Palau, Panama, Papua New Guinea, Paraguay, Peru, the Philippines, Puerto Rico, Rwanda, Senegal, Seychelles, Sierra Leone, Singapore, South Africa, South Korea, Sri Lanka, Tanzania, Tonga, Trinidad and Tobago, Trkiye, the U.S. Virgin Islands, Uganda, the United States, Uruguay, Venezuela, Zambia, and Unveiled at Google's I/O 2025 developer conference earlier this week, Veo 3 introduces several upgrades over its predecessor, Veo 2. Built on a combination of advanced AI systems — natural language processing, text-to-video diffusion, audio synthesis, and lip-syncing tech — what makes this version particularly interesting is its ability to generate videos with audio for the first time. Google says Veo 3 can interpret detailed prompts — everything from moods and tones to cultural settings — and bring them to life with cinematic alongside its promise, Veo 3 also raises familiar concerns about misuse. The ease with which users can now create convincing fake interviews, protest clips, or even fabricated news segments highlights the growing challenge of verifying visual content online.
Time of India
24-05-2025
- Entertainment
- Time of India
Google's Veo 3 videos go viral: From stand-up comedy to Will Smith eating spaghetti, netizens face an existential crisis
Google announced an impressive number of artificial intelligence (AI) innovations at I/O 2025 on Tuesday, May 20, 2025. With the new AI innovations, they made it clearer than that how powerful the AI firm has become. Not only this, but Veo 3, the new content generation, Google is turning the wild ideas into jaw-dropping videos with nothing but a few prompts and a dash of digital magic. However, these kinds of innovations are nothing new in the AI space, and there is a slew of impressive AI-generated content, including realistic photos, incredible music, and videos that almost look professionally shot. But don't be shocked if we say that something more interesting and promising is being launched in the AI space. Yes, some videos created by Veo 3 are going viral on social media and have almost broken the internet. Say goodbye to the silent era of video generation: Introducing Veo 3 — with native audio generation. 🗣️Quality is up from Veo 2, and now you can add dialogue between characters, sound effects and background noise. Veo 3 is available now in the @GeminiApp for Google AI Ultra… Veo 3 doesn't just generate video; it creates audio as well Since the day Google announced its AI innovations, social media has been flooded with videos made with Veo 3 that netizens are in denial about believing. Among all the videos, one video is of a man doing a stand-up comedy show in a small venue and telling a joke, which is actually a prompt. (However, the authenticity of the video is not confirmed, shared by an X user @MrBahaudin.) Well, the video quickly went viral, and netizens are stunned by it. NO WAY. It did it. And, was that, actually funny?Prompt:> a man doing stand up comedy in a small venue tells a joke (include the joke in the dialogue)#veo3 Well, this is not the only one; there is another where actor Will Smith is seen eating spaghetti, which looks as real as it can. Just got access to Veo 3 and the first thing I did was try the Will Smith spaghetti test. SOUND ON In these videos, Veo 3 has created audio to match the action in the scenes, including the background noise and dialogue. It indicates that it did a great job of maintaining character, location, and object consistency from scene to scene. Veo 3's AI prompts video stuns the internet Well, there are plenty of such clips, which have blown the minds of netizens on the internet. One of the videos that surfaced on Instagram's page called Evolving AI is a compilation of all the videos by @ It is a hilarious and thought-provoking video. Created by using Google's Veo 3 AI video model, AI-generated characters challenge the very idea of their existence. Netizens looked cooked in the comment section, and one said, "We are cooked 😄 well done Google!" "Future generations won't believe real historical events ever happened, even if it was filmed. The universal truths that we all agree on will be no more. Nothing in history nor the present will ever be certain again," another commented. "5⭐️Better than 9 out of 10 "made for Netflix" films. But, maybe that's just the prompt talking ." "I feel insecure and starting to have existential questions" "This is beyond trippy man, can you imagine five years from now??? It'll be a completely different world" While one said, "You can tell this is fake because you feel nothing. No real human emotions coming thru these clips." To stay updated on the stories that are going viral, follow Indiatimes Trending.
Express Tribune
23-05-2025
- Entertainment
- Express Tribune
The prompt theory
Listen to article Are we living in the most oppressive timeline or the most fascinating one? It is hard to tell. One minute, technological breakthroughs take your breath away; the next, mankind does something so stupid you start worrying about its future. Before introducing the technological bit, let me inform you that this piece is not about tech development - nor the fascinating twist to which it owes its title. In this climate, it is hard to talk about anything other than the India-Pakistan conflict. But technology provides an interesting trigger. You will see. Last week, Google launched Veo 3, its most advanced AI video generator. Its biggest highlight is that it creates videos with synchronised audio, including dialogue, sound effects and music. This is a significant step-up from Veo 2 and OpenAI's Sora, which lacked native audio and had lower overall quality and control. Seeing is believing, and it is the kind of technology that can instantly kill the entire entertainment industry. Forget about actors, directors, special effects artists and musicians. Just feed text and pictures relating to a script through a prompt and watch it come true. If, for some reason, you feel that the outcome is subpar, take solace in the fact that the previous model, Veo 2, was released only six months ago. And who knows what Sora and other AI brands will come up with next. What intrigued me the most while going through various generated videos was a series of clips someone with remarkable humour and a sense of irony had produced. The realistic characters in those clips can be heard and seen rejecting what they call the "prompt theory" — the idea that they are created by prompts and a random prompt writer is deciding their fate. If this rings a bell, that is because it is a play on the idea many know as the simulation theory — the belief that we exist in an elaborate four-dimensional realistic simulation. You will be surprised to know how many influential figures believe in that. My response to the notion is simple. Call it by any name, life is life. As long as we can live and feel it, it is life. But here's the rub. We think we are free agents. But the truth is, what we call free will is also an algorithm. A complex, partly biological, partly sociologically informed algorithm, I grant you — but an algorithm nonetheless. So, call it predestined or whatever you will; we are creatures of our circumstances. Now, from the sublime to the stupid. India says its war on Pakistan is only suspended, not ended. I wish the Indian media could also suspend its war on the sensibilities of its viewers — and common sense. But the Punch and Judy show continues unabated. Remember that dark night when overzealous Indian anchors were claiming that their country's forces had destroyed Pakistan? Once their unprofessionalism was badly exposed, not everyone showed contrition. One anchor recorded a video and shared it on social media. In the clip, he asked what was so wrong in lying about Pakistan. His video was also shared by the ruling BJP's official handle. Another genius claimed there was no point in getting alarmed by this as it was just an average, everyday psychological operation. How is it a psychological operation if the end user is your domestic audience? He did not care to elaborate. The impunity with which these elements can carry on without consequences should tell you you are dealing with an exceptional condition. It is one sign, at any rate. But if that was the situation within India, guess what was happening outside. The Indian diaspora is known for its competence. Many have well-deservedly risen to the top in their respective professions. Being sympathetic to the country of your origin is not unheard of - especially when it is a cultural powerhouse like India. So far, so good. But until this government came to power, the Indian state either desisted from exploiting their loyalty or used it subtly and sparingly. However, this government, since 2014, has leaned heavily on them as well as native Indophiles, shunning those who refused to play ball - particularly in media and academia. So, those who did comply have risen in the ranks. And their spin is not subtle. Consider the tweet by Shashank Joshi, the Defence Editor of The Economist, commenting on Jim Sciutto's admirable story citing French authorities: "France acknowledges the loss of an Indian Rafale. According to the images circulating, which appear to be original, it may have been the first of 36 Rafale that entered service in the Indian Air Force. The next question is whether it was hit by a Chinese PL-15 missile." Do you see the sleight of hand? Indian Rafale, Chinese PL-15? China-India dispute framing tracks better in the West. Otherwise, if PL-15s are Chinese, Rafales are French. If Rafales are Indian, PL-15s are Pakistani. But framing it as a China-India conflict says India is fighting the "good fight" against China; Pakistan is just in the way - only collateral damage. In his tweet dated May 10, some serious foreshadowing: "I think there will be a lot of interest in satellite images in the coming days to sort through which side's claims of striking which targets are true and which are exaggerated or false, particularly in relation to strikes on air defence systems, air bases and other military sites." No kidding. The New York Times published a story on May 14 titled, "India and Pakistan Talked Big, but Satellite Imagery Shows Limited Damage." The story mentioned a May 10 claim by Pakistani authorities of damaging the Udhampur air base and produced a satellite image taken two days later - allowing two whole days for cover-up and repair. The Pakistani images are undated. Why is this suspicious? Because, to the current Indian government, image is more important than reality. Another gentleman torn from the same cloth: Ravi Agrawal, Foreign Policy's Editor-in-Chief, who appeared on CNN's marathon transmission presenting Indian strikes as a rational choice. The comfort with which this lot brushes aside the questions about inculpatory evidence is breathtaking. But then there were researchers from the diaspora who wrote research papers trying to scientifically validate PM Modi's belief that cloud cover can help evade radar detection. This is some next-level mind-bending spin. And here we have our dear Pakistani colleagues who are beating themselves up just because, in view of their mild-mannered and responsible coverage, their government praised their behaviour. When you are fighting an existential battle against such industrial-scale falsehoods, what do you care who praises you? Tell yourselves you are better at civilisational prompt writing.
Time of India
23-05-2025
- Entertainment
- Time of India
Google unveils Veo 3, an AI-powered video generation tool: Check price, features, who can use Veo 3, and other details
Google Veo 3 In a historic announcement at Google I/O 2025, the tech giant unveiled its most ambitious AI video generation model to date: Veo 3. This latest iteration marks a significant leap forward in artificial intelligence, blurring the boundaries between human-created and machine-generated content. What sets Veo 3 apart is not just its ability to create strikingly realistic visuals from simple text prompts, but its newly introduced capability to generate synchronized audio—dialogue, background noise, music, and ambient sound—all in one seamless output. For decades, the dream of creating cinematic experiences through AI has remained largely in the realm of science fiction. Today, Veo 3 transforms that dream into a tangible reality. Designed for content creators, educators, marketers, and filmmakers, Veo 3 dramatically lowers the barrier to professional-grade video production. With just a few words, users can generate rich, immersive videos that would have previously required a crew, equipment, and significant post-production resources. The implications are immense, not only for entertainment but for education, journalism, business, and beyond. What is Google Veo 3 and why it matters Google's Veo 3 is the third and most advanced version of its generative video model. Unlike its predecessor, Veo 2, which could only create silent video clips, Veo 3 adds the missing piece: natural-sounding, context-aware audio. This includes: by Taboola by Taboola Sponsored Links Sponsored Links Promoted Links Promoted Links You May Like The Killer New Toyota 4Runner Is Utter Perfection (Take A Look) MorninJoy Undo Synchronized voiceovers Emotionally-matched dialogue Authentic sound effects (e.g., footsteps, background chatter) Musical accompaniments aligned with the scene's tone and pacing This fusion of sound and vision results in an experience that is eerily close to real life. One of Veo 3's key differentiators is its ability to synchronize lip movements with dialogue, making characters in the video appear convincingly human. Furthermore, it understands context. For example, if a user inputs the prompt: "a thunderstorm at sea with a ship struggling against the waves," the result is a cinematic video complete with storm sounds, creaking wood, and urgent narration—entirely AI-generated. How Veo 3 works: The tech behind the magic Veo 3 is built upon a foundation of multimodal AI, combining natural language processing (NLP), text-to-video diffusion models, and text-to-speech synthesis with generative adversarial networks (GANs). Key features include: Text-to-video translation : Converts complex prompts into coherent scene sequences with realistic motion and object physics. Audio rendering layer : Uses AI voice models and sound synthesis to create environment-appropriate audio. Lip synchronization engine : Matches generated speech with facial movements using motion prediction algorithms. Temporal consistency engine : Ensures frame-by-frame continuity and smooth transitions in animations. Google's use of its Gemini Ultra foundation model also enables Veo 3 to understand nuanced instructions such as tone of voice, cinematic mood, or specific cultural settings. How creators are using Veo 3 Since its debut, creators have flocked to Veo 3 to explore its capabilities. Viral content quickly surfaced across social platforms like X (formerly Twitter), YouTube, and TikTok. Stand-up comedy video : One viral video featured a completely AI-generated stand-up routine, with not only a virtual comedian on stage but also background audience laughter and responsive timing. No cameras. No mics. Just a text prompt. Historical reenactment : Another clip depicted Pythagoras explaining his theorem. The video included historically accurate attire, an ancient Greco-Roman setting, and narrated explanations—impressively detailed and educational. Music video generation : One user created a full music video, from lyrics and beat to visuals and dance choreography. The harmony between video cuts and music rhythm amazed many viewers and raised the bar for indie production. Who can use Veo 3 and how: Check how to access and know its pricing As of May 2025, Veo 3 is available exclusively in the United States and only for premium subscribers. Access is granted through: Platform: Google Gemini App and Flow Service tier: Gemini Ultra Monthly subscription: $249.99 It's also integrated into Google's Vertex AI suite, making it available for enterprise-level customers, media studios, and advertising agencies. While this price point is clearly aimed at serious professionals, Google has hinted at future pricing models that could allow broader access, especially as demand scales. Why Veo 3 could change everything What makes Veo 3 more than just a tool is the democratization of creativity it enables. For decades, creating even a short professional video required expensive equipment, a team of specialists, and post-production work. With Veo 3, creators now just need an idea and a few sentences. This shift redefines how we approach storytelling. Students can create history projects that look like documentaries. Small businesses can produce polished ads without agencies. Independent filmmakers can prototype entire scenes before investing in production. Google also touts Veo 3's educational potential, especially in multilingual regions. The model can render the same video in different languages with native-style voiceovers, offering powerful tools for global teaching and accessibility. When will Veo 3 come to India and other countries Currently, there is no confirmed timeline for Veo 3's global rollout, including availability in India. However, given the country's booming content creation economy and rising adoption of generative AI, industry watchers expect India to be among the first wave of international markets. In the meantime, Google is working to expand infrastructure and compliance for its Vertex AI and Gemini platforms in Asia. Localization support, including regional languages, could be a key part of Veo 3's expansion strategy. Veo 3 and the deepfake dilemma: How safe is too safe As with any powerful AI tool, Veo 3 raises questions around: Deepfake misuse Content authenticity Intellectual property rights Bias in voice and character generation Google claims to have embedded robust watermarking and usage detection systems to combat misuse. Additionally, all content generated with Veo 3 includes metadata tags for AI attribution. Still, ongoing discussions about ethics and regulation are likely to follow Veo's broader adoption. Google Veo 3 related FAQs What is Google Veo 3 and how is it different from older versions? Veo 3 is Google's AI video model that now includes synchronized audio, unlike Veo 2 which only produced silent visuals. How can I access Google Veo 3 and what does it cost? It is currently available in the U.S. via the Gemini app's Ultra plan for $249.99/month and through Vertex AI for enterprise users. Can Veo 3 replace human filmmakers? Not entirely. While Veo 3 is powerful, it serves as a tool for creative augmentation, not a total replacement for human storytelling, direction, or emotion. When will Veo 3 launch in India? No official date yet, but Google is expected to expand to India soon, especially with high creator interest. AI Masterclass for Students. Upskill Young Ones Today!– Join Now
