Forwarded from Mishaal's Android News Feed
Gemini for Android is getting upgraded with new contextual capabilities!
- A new “ask this video” chip will appear if you invoke Gemini while a YouTube video is open. Gemini will then use the YouTube video as context.
- A new “ask this PDF” chip will appear when you’re viewing a PDF file in supported apps. Tapping this will let you ask questions about/summarize the document. This requires Gemini Advanced, though, because of its longer context window.
- You’ll be able to drag and drop generated images directly into Gmail, Messages, and other supported apps.
- Finally, devices with the next-gen multimodal Gemini Nano will surface dynamic suggestions. The way this works is that a screenshot of the screen is fed to Gemini Nano, which can then generate chips of potential things you might want to ask about.
- A new “ask this video” chip will appear if you invoke Gemini while a YouTube video is open. Gemini will then use the YouTube video as context.
- A new “ask this PDF” chip will appear when you’re viewing a PDF file in supported apps. Tapping this will let you ask questions about/summarize the document. This requires Gemini Advanced, though, because of its longer context window.
- You’ll be able to drag and drop generated images directly into Gmail, Messages, and other supported apps.
- Finally, devices with the next-gen multimodal Gemini Nano will surface dynamic suggestions. The way this works is that a screenshot of the screen is fed to Gemini Nano, which can then generate chips of potential things you might want to ask about.
👍20
Forwarded from Mishaal's Android News Feed
This media is not supported in your browser
VIEW IN TELEGRAM
Circle to Search can now help students do their homework! If you select a math or physics word problem using Circle to Search, Google will now generate step-by-step instructions on how to solve the problem, without leaving the page you’re on.
Later this year, Circle to Search will be able to solve more complex math problems that involve symbolic formulas, diagrams, graphs, and more thanks to LearnLM, Google’s new family of language models fine-tuned for learning.
Later this year, Circle to Search will be able to solve more complex math problems that involve symbolic formulas, diagrams, graphs, and more thanks to LearnLM, Google’s new family of language models fine-tuned for learning.
👍35👎9❤4
Forwarded from Mishaal's Android News Feed
Here's a quick demo of the new, more context-aware Gemini app for Android.
The new app supports drag-and-drop of images, can show take a YouTube video as context if you tap "Ask this video", and can take a PDF as context if you tap "Ask this PDF".
The new app supports drag-and-drop of images, can show take a YouTube video as context if you tap "Ask this video", and can take a PDF as context if you tap "Ask this PDF".
YouTube
Demo of the new Gemini app for Android
A demo of the updated Gemini app for Android announced at Google I/O 2024.
Forwarded from ATT • Tech News (aanjaneya_meaow)
OpenAI is considering allowing generation of NSFW content
OpenAI is considering allowing NSFW content generation on DALL-E and more importantly, ChatGPT. Despite this, they will still have a ban on deepfake content, especially adult related. This suggestion was revealed in a recent document that talks about the development of the AI.
Now, the main point is: OpenAI MAY allow NSFW soon. And as it's the leader in AI space, consider how other AI services (ESPECIALLY Character.ai) react.
🔗 Link
🧑💻 @agamtechtricks
OpenAI is considering allowing NSFW content generation on DALL-E and more importantly, ChatGPT. Despite this, they will still have a ban on deepfake content, especially adult related. This suggestion was revealed in a recent document that talks about the development of the AI.
Now, the main point is: OpenAI MAY allow NSFW soon. And as it's the leader in AI space, consider how other AI services (ESPECIALLY Character.ai) react.
Please open Telegram to view this post
VIEW IN TELEGRAM
👍26🤣20👎12💘4🙈3🤔1
Chinese Smartphones Brands Cannot AI Erase China's President Image and Tiananmen Square
Source
Chinese smartphones brands like Oppo, Xiaomi, Vivo, Huawei etc and other cannot remove images of China's President Xi Jinping and the historical place Tiananmen Square using their AI Erase tools
Vivo, Huawei AI erase tools simply says that "the image contains sensitive information", thus declining to remove the content
Realme or Oppo smartphones AI tools gives warning of "high-risk content", thus failing to edit the image 💀
However, Chinese smartphones brands allows to AI erase images of other world political leaders like Russia's President Vladimir Putin easily
In global ROMs also, Xiaomi allows to remove or edit images of China's President Xi Jinping but you cannot do the same in China ROMs
Follow @TechLeaksZone
Source
Chinese smartphones brands like Oppo, Xiaomi, Vivo, Huawei etc and other cannot remove images of China's President Xi Jinping and the historical place Tiananmen Square using their AI Erase tools
Vivo, Huawei AI erase tools simply says that "the image contains sensitive information", thus declining to remove the content
Realme or Oppo smartphones AI tools gives warning of "high-risk content", thus failing to edit the image 💀
However, Chinese smartphones brands allows to AI erase images of other world political leaders like Russia's President Vladimir Putin easily
In global ROMs also, Xiaomi allows to remove or edit images of China's President Xi Jinping but you cannot do the same in China ROMs
Follow @TechLeaksZone
🤣135🤡9👍8🤨1
Forwarded from ATT • Tech News (Agam)
Google Photos ‘Ask Photos’ will let you search your images with voice and text prompts
"Ask Photos" will make it easier to search across your photos and videos with the help of Gemini models. It goes beyond simple search to understand context and answer more complex questions
🔗 9to5Google
🧑💻 @agamtechtricks
"Ask Photos" will make it easier to search across your photos and videos with the help of Gemini models. It goes beyond simple search to understand context and answer more complex questions
Please open Telegram to view this post
VIEW IN TELEGRAM
👍20🆒4❤1👏1
Forwarded from ATT • Tech News (Agam)
NotebookLM gains podcast-like audio overviews with live Q&A
Google is upgrading NotebookLM, for asking questions about a set of documents, with new “audio overviews” that generate a podcast-like discussion between two speakers.
This feature was presented from the perspective of someone who learns best when listening rather than just reading. In the on-stage example, NotebookLM was given a set of simple physics lessons as its base context. By generating an audio overview, the tool generated a conversation between two speakers and voiced both participants in real time.
🔗 9to5Google
🧑💻 @agamtechtricks
Google is upgrading NotebookLM, for asking questions about a set of documents, with new “audio overviews” that generate a podcast-like discussion between two speakers.
This feature was presented from the perspective of someone who learns best when listening rather than just reading. In the on-stage example, NotebookLM was given a set of simple physics lessons as its base context. By generating an audio overview, the tool generated a conversation between two speakers and voiced both participants in real time.
Please open Telegram to view this post
VIEW IN TELEGRAM
❤8👍4👌1🤣1
Forwarded from ATT • Tech News (Agam)
Gemini Live
In Gemini Live, you’ll get a fullscreen experience with a cool audio waveform effect. This will let you have a 2-way dialogue, with Gemini returning concise responses.
You can speak at your own pace, with Google adapting, and interrupt Gemini as it’s replying to add new information or ask for clarification.
Available for Gemini Advanced subscribers, it’s launching in the coming months.
🔗 9to5Google
🧑💻 @agamtechtricks
In Gemini Live, you’ll get a fullscreen experience with a cool audio waveform effect. This will let you have a 2-way dialogue, with Gemini returning concise responses.
You can speak at your own pace, with Google adapting, and interrupt Gemini as it’s replying to add new information or ask for clarification.
Available for Gemini Advanced subscribers, it’s launching in the coming months.
Please open Telegram to view this post
VIEW IN TELEGRAM
Please open Telegram to view this post
VIEW IN TELEGRAM
👍22🤣2❤1
Forwarded from ATT • Tech News (Agam)
This media is not supported in your browser
VIEW IN TELEGRAM
Project Astra
The Astra demo Google showed — single-take in real-time — pointed a phone at objects as someone issued commands or questions, with Gemini recognizing what’s in front of it in near real-time. You can show it a cityscape and ask what neighborhood you’re in, or inquire about code.
This is built on the Gemini 1.5 Pro mode and “other task specific models.” Google says it’s “designed to process information faster by continuously encoding video frames,” with reducing the response times to “something conversational” a “difficult engineering challenge.”
🔗 9to5Google
🧑💻 @agamtechtricks
The Astra demo Google showed — single-take in real-time — pointed a phone at objects as someone issued commands or questions, with Gemini recognizing what’s in front of it in near real-time. You can show it a cityscape and ask what neighborhood you’re in, or inquire about code.
This is built on the Gemini 1.5 Pro mode and “other task specific models.” Google says it’s “designed to process information faster by continuously encoding video frames,” with reducing the response times to “something conversational” a “difficult engineering challenge.”
…combining the video and speech input into a timeline of events, and caching this information for efficient recall.”
Please open Telegram to view this post
VIEW IN TELEGRAM
🆒40👍6🤣5😭4
Forwarded from ATT • Tech News (Agam)
Imagen 3
Imagen 3 is positioned as Google’s “highest-quality” text-to-image model and offers a few improvements over the Imagen 2 model we’ve seen in Gemini and Bard.
Imagen 3 is said to bring a higher level of detail in images without as many visual artifacts and impurities in generated images. The images are more photorealistic and lifelike when requested.
🔗 9to5Google
🧑💻 @agamtechtricks
Imagen 3 is positioned as Google’s “highest-quality” text-to-image model and offers a few improvements over the Imagen 2 model we’ve seen in Gemini and Bard.
Imagen 3 is said to bring a higher level of detail in images without as many visual artifacts and impurities in generated images. The images are more photorealistic and lifelike when requested.
Please open Telegram to view this post
VIEW IN TELEGRAM
❤17👍6🤣2
Forwarded from ATT • Tech News (Agam)
Google launches ‘Veo,’ an AI video generation tool
This model is built specifically for video generation that understands visual semantics and natural language, similar to other modern models. That approach brought into video generation offers results that can be creatively tailored to fit certain styles.
Google notes that the Veo model will be able to understand “cinematic terms” in the user’s prompts, like aerial shots and timelapse formats. Veo is capable of generating videos in 1080p that can last beyond a minute, which surpasses current models like OpenAI’s Sora, maxing out at 60 seconds.
🔗 9to5Google
🧑💻 @agamtechtricks
This model is built specifically for video generation that understands visual semantics and natural language, similar to other modern models. That approach brought into video generation offers results that can be creatively tailored to fit certain styles.
Google notes that the Veo model will be able to understand “cinematic terms” in the user’s prompts, like aerial shots and timelapse formats. Veo is capable of generating videos in 1080p that can last beyond a minute, which surpasses current models like OpenAI’s Sora, maxing out at 60 seconds.
Please open Telegram to view this post
VIEW IN TELEGRAM
Please open Telegram to view this post
VIEW IN TELEGRAM
❤🔥17🤣8👍7🆒3
Forwarded from ATT • Tech News (Agam)
Here's a list of all the new features and updates in the Google I/O 2024 Event:
• Google Photos' "Ask Photos"
• NotebookLM's Audio Overviews
• New Model: Gemini Nano
• New Gemini Upgrades
• Gemini Live
• Project Astra
• More Gemini Updates
• Gmail on Phone Updates
• Gemini 1.5 Pro-powered side panel
• Gemini in Android Studio
• Gemini Nano in Chrome
• New Google Play features
• Veo - AI Video Generation Tool
• Imagen 3
• New 'Web' filter in Google Search as AI answers can't be disabled
Exclusive content, only at Agam's Tech Tricks 🗣🔥
Credits for the individual posts: 9to5Google, Mishaal Rahman
👨💻 @agamtechtricks
• Google Photos' "Ask Photos"
• NotebookLM's Audio Overviews
• New Model: Gemini Nano
• New Gemini Upgrades
• Gemini Live
• Project Astra
• More Gemini Updates
• Gmail on Phone Updates
• Gemini 1.5 Pro-powered side panel
• Gemini in Android Studio
• Gemini Nano in Chrome
• New Google Play features
• Veo - AI Video Generation Tool
• Imagen 3
• New 'Web' filter in Google Search as AI answers can't be disabled
Exclusive content, only at Agam's Tech Tricks 🗣🔥
Credits for the individual posts: 9to5Google, Mishaal Rahman
Please open Telegram to view this post
VIEW IN TELEGRAM
❤🔥16👍1
Google I/O AI Count
During the 110mins Google I/O event, Google speakers spoke the 'AI' word 121 times
It's surprisingly less than the Google I/O 2023 event, where the 'AI' word was spoken 137 times
During the Samsung Galaxy Unpacked Event 2024, the 'AI' was spoken just 56 times
Next up: AI word count at Apple's WWDC 2024 event
Follow @TechLeaksZone
During the 110mins Google I/O event, Google speakers spoke the 'AI' word 121 times
It's surprisingly less than the Google I/O 2023 event, where the 'AI' word was spoken 137 times
During the Samsung Galaxy Unpacked Event 2024, the 'AI' was spoken just 56 times
Next up: AI word count at Apple's WWDC 2024 event
Follow @TechLeaksZone
😁44🎃3👍1
Apple's iOS 17.5 is reintroducing images that users deleted years ago
Reddit
A bug in iOS 17.5 is resurfacing photos that iPhone users deleted years ago in the Photos app as the most recent images
"When in conversation with my partner, I went to send a picture and saw that the latest pictures were NSFW material we’d made years ago when we were living apart. It was permanently deleted years ago, but magically it’s back?"
"I checked my iPad and it also has pictures (some art work I did years ago). I feel so uncomfortable."
"Same here. I have 4 pics from 2010 that keep reappearing as the latest pics uploaded to iCloud. I have deleted them repeatedly,” says another user
Apple hasn’t commented on the problem, but it raises questions about privacy and data retention
Follow @TechLeaksZone
A bug in iOS 17.5 is resurfacing photos that iPhone users deleted years ago in the Photos app as the most recent images
"When in conversation with my partner, I went to send a picture and saw that the latest pictures were NSFW material we’d made years ago when we were living apart. It was permanently deleted years ago, but magically it’s back?"
"I checked my iPad and it also has pictures (some art work I did years ago). I feel so uncomfortable."
"Same here. I have 4 pics from 2010 that keep reappearing as the latest pics uploaded to iCloud. I have deleted them repeatedly,” says another user
Apple hasn’t commented on the problem, but it raises questions about privacy and data retention
Follow @TechLeaksZone
🤡73🤣29👍6👀3❤2😡1
Samsung launched its most expensive MicroLED TV in South Korea
Samsung has launched a 114-inch version of its Micro LED TV in South Korea on popular demand (likely mostly from millionaires)
It had 89-inch and 101-inch sizes earlier, and adding the new size will “broaden the range of consumer choices.”
The 114-inch version of Samsung's Micro LED TV costs a whopping KRW 180 million ($132,630 or INR 1,10,59,408) 💀
Since its so much expensive, Samsung is giving away an 85-inch Neo QLED TV (QND900) for FREE and JBL L100 MK2 speaker at a discount.
Not only this, Samsung is also offering a KRW 3 million voucher (~ $2215 or INR 1,84,847) to stay at the Signiel Seoul luxury hotel to those who buy the 114-inch Micro LED TV this month.
Follow @TechLeaksZone
Samsung has launched a 114-inch version of its Micro LED TV in South Korea on popular demand (likely mostly from millionaires)
It had 89-inch and 101-inch sizes earlier, and adding the new size will “broaden the range of consumer choices.”
The 114-inch version of Samsung's Micro LED TV costs a whopping KRW 180 million ($132,630 or INR 1,10,59,408) 💀
Since its so much expensive, Samsung is giving away an 85-inch Neo QLED TV (QND900) for FREE and JBL L100 MK2 speaker at a discount.
Not only this, Samsung is also offering a KRW 3 million voucher (~ $2215 or INR 1,84,847) to stay at the Signiel Seoul luxury hotel to those who buy the 114-inch Micro LED TV this month.
Follow @TechLeaksZone
🌚34🤡16⚡5🤣5👍2❤1