正在加载视频...

视频加载失败

加载此视频时出现问题。这可能是由于临时网络问题，或视频可能不可用。

🐕🐈 Chinese developers created an AI collar that “translates” barking and meowing The gadget runs on Alibaba Cloud’s Qwen model. Developers trained the neural network on 1.5 million animal sound recordings and thousands of hours of cat and dog videos. The creators claim translation accuracy of up to 94.6%... show more

NEXTA

1,318,640 subscribers

33,209 次观看 • 3 个月前 •via X (Twitter)

健康养生科学技术动物

Anya Rossi• Live Now

Private livecam show

0 条评论

暂无评论

原始帖子的评论将显示在这里

相关视频

🐶 A Chinese startup built a collar that claims to translate your dog's barks into human words in real time. AI reads each sound, matches it against a million recordings, and guesses what your pet feels. 📱 The catch: the accuracy claims aren't independently proven yet. ✨ #omgfacts #Dogs #AI

🐶 A Chinese startup built a collar that claims to translate your dog's barks into human words in real time. AI reads each sound, matches it against a million recordings, and guesses what your pet feels. 📱 The catch: the accuracy claims aren't independently proven yet. ✨ #omgfacts #Dogs #AI

OMGFacts

16,450 次观看 • 1 个月前

Turkish Leftist Twitch streamer Hasan Piker appeared to use a shock collar on his dog as it walked out of frame during a livestream, causing the dog to yelp in pain and cower away from Piker. Many have suggested that Piker is using the dog as a prop for his streams, hence using a shock collar to prevent the dog from moving during the recordings. Piker admitted using a shock collar on his dog in prior streams, describing it as "incredibly effective." In one stream, a guest noted that the shock collar was "too tight.” Piker appeared to comment on the collar itself before pausing and saying the collar was tight because of “neck meat.” Follow: AF Post

Turkish Leftist Twitch streamer Hasan Piker appeared to use a shock collar on his dog as it walked out of frame during a livestream, causing the dog to yelp in pain and cower away from Piker. Many have suggested that Piker is using the dog as a prop for his streams, hence using a shock collar to prevent the dog from moving during the recordings. Piker admitted using a shock collar on his dog in prior streams, describing it as "incredibly effective." In one stream, a guest noted that the shock collar was "too tight.” Piker appeared to comment on the collar itself before pausing and saying the collar was tight because of “neck meat.” Follow: AF Post

AF Post

116,883 次观看 • 11 个月前

Tell Mao Ning 毛宁 that the rest of the world does not tolerate animal cruelty like what happened to #WangWang and her pups. China 🇨🇳 has an obsession with being cruel to cats and dogs and hamsters and sharing the videos on regular Chinese social media! Chinese even show off animal cruelty on X, with threatening messages to Americans and anyone against animal abuse!

Tell Mao Ning 毛宁 that the rest of the world does not tolerate animal cruelty like what happened to #WangWang and her pups. China 🇨🇳 has an obsession with being cruel to cats and dogs and hamsters and sharing the videos on regular Chinese social media! Chinese even show off animal cruelty on X, with threatening messages to Americans and anyone against animal abuse!

Phaedra

12,967 次观看 • 1 个月前

Beginning June 21st, the dog and cat meat trade intensifies in China’s city of Yulin. Truckloads of dogs and cats will be driven long distances in horrific conditions to be brutally slaughtered for their meat. Please urge China to end the animal welfare and public health nightmare of Yulin’s dog and cat meat trade:

Beginning June 21st, the dog and cat meat trade intensifies in China’s city of Yulin. Truckloads of dogs and cats will be driven long distances in horrific conditions to be brutally slaughtered for their meat. Please urge China to end the animal welfare and public health nightmare of Yulin’s dog and cat meat trade:

Humane World for Animals

10,243 次观看 • 3 个月前

A cat meat shop in Qingyuan, Guangdong, #China Nearly 100 cats were killed alive on the street. Food safety cannot be guaranteed. Dogs and cats are not animals meant for human consumption. Say no to cat and dog meat — support the inclusion of animal protection in legislation. #StopChinaCatTorture

A cat meat shop in Qingyuan, Guangdong, #China Nearly 100 cats were killed alive on the street. Food safety cannot be guaranteed. Dogs and cats are not animals meant for human consumption. Say no to cat and dog meat — support the inclusion of animal protection in legislation. #StopChinaCatTorture

ALEJOCat_Sp

10,809 次观看 • 1 年前

🎊 Kudos to the teams at Qwen on the launch of Qwen 3.5 with Qwen3.5-397B-A17B. 🙌 Developers can start building today for free: Or download and customize it with NVIDIA NeMo:

🎊 Kudos to the teams at Qwen on the launch of Qwen 3.5 with Qwen3.5-397B-A17B. 🙌 Developers can start building today for free: Or download and customize it with NVIDIA NeMo:

NVIDIA AI Developer

26,904 次观看 • 7 个月前

The most downloaded AI on earth is now Chinese. Alibaba just gave away a model that matches Claude's flagship, and it literally runs on a $700 used graphics card. The Qwen models crossed 3 BILLION downloads in six months. Hugging Face counted 418 million downloads for Google this year, and 227 million for Meta. Alibaba cleared more than four times both of them combined. Then today it released Qwen3.8-27B under an Apache 2.0 license. The model has 27 billion parameters, native vision, and a 262,000 token context window. Developers are running it locally on 17 gigabytes of memory, on used cards that cost a few hundred dollars. Alibaba's own benchmark table claims it beats Opus 4.6 Max on computer use by 84.3 to 72.7, on mobile use by 81.9 to 62, and on visual math by 94.6 to 65.5. Those numbers come from the vendor and nobody has independently verified them yet, so treat them as a claim. But the generation over generation jumps are harder to wave away: On DeepSWE the score went from 13.3 to 42.2. On software engineering it went from 49.3 to 79.0. That happened in ONE release cycle. And Apache 2.0 means anyone can download the weights, modify them, build products on them, sell those products, and never pay or ask permission. It cannot be revoked. Once the file is on your drive it is yours permanently. 3 billion downloads means those files already sit on machines in every country on Earth. Alibaba could delete everything tomorrow and it would change nothing. Washington spent 4 years building an export control regime around chips, model weights, and entity lists. Every piece of it assumes a chokepoint exists somewhere. A fab, a shipment, a company that can be told no. But there is no chokepoint for a file that has already been copied three billion times. And the copying compounds. Hugging Face counted 151,448 models built on top of Qwen, which is 2.6x Meta's entire footprint and 4.7x the number of Llama repositories. New ones appear at roughly 200 a day. The report says Qwen has become "part of the default workflow for developers deciding what models to fine-tune and deploy." Alibaba is also pushing Qwen through its cloud into Southeast Asia and Africa, markets where American labs have almost no presence, and where a very large share of the next generation of developers will learn to build. Meta and Nvidia have both rushed out new open models in recent weeks. That is what a response looks like when you feel the floor move. And to be clear, these are download and derivative numbers, not usage numbers. ChatGPT and Claude cannot be downloaded at all, so they do not appear in this comparison. What the figures measure is what developers choose to build on top of, which is a different question from what consumers type into a box. That is also why it matters MORE. Consumer habits change in an afternoon. Infrastructure choices last a decade, because everything built on top has to be rewritten to undo them. The American labs are valued on an assumption that frontier intelligence stays scarce, expensive, and rented by the token. Alibaba just made a version of it free, permanent, and small enough to run on hardware people already own. You will not get an announcement when the software you use every day starts running on a Chinese model underneath. Go and count how many of the tools you rely on could be rebuilt on free weights this year.

The most downloaded AI on earth is now Chinese. Alibaba just gave away a model that matches Claude's flagship, and it literally runs on a $700 used graphics card. The Qwen models crossed 3 BILLION downloads in six months. Hugging Face counted 418 million downloads for Google this year, and 227 million for Meta. Alibaba cleared more than four times both of them combined. Then today it released Qwen3.8-27B under an Apache 2.0 license. The model has 27 billion parameters, native vision, and a 262,000 token context window. Developers are running it locally on 17 gigabytes of memory, on used cards that cost a few hundred dollars. Alibaba's own benchmark table claims it beats Opus 4.6 Max on computer use by 84.3 to 72.7, on mobile use by 81.9 to 62, and on visual math by 94.6 to 65.5. Those numbers come from the vendor and nobody has independently verified them yet, so treat them as a claim. But the generation over generation jumps are harder to wave away: On DeepSWE the score went from 13.3 to 42.2. On software engineering it went from 49.3 to 79.0. That happened in ONE release cycle. And Apache 2.0 means anyone can download the weights, modify them, build products on them, sell those products, and never pay or ask permission. It cannot be revoked. Once the file is on your drive it is yours permanently. 3 billion downloads means those files already sit on machines in every country on Earth. Alibaba could delete everything tomorrow and it would change nothing. Washington spent 4 years building an export control regime around chips, model weights, and entity lists. Every piece of it assumes a chokepoint exists somewhere. A fab, a shipment, a company that can be told no. But there is no chokepoint for a file that has already been copied three billion times. And the copying compounds. Hugging Face counted 151,448 models built on top of Qwen, which is 2.6x Meta's entire footprint and 4.7x the number of Llama repositories. New ones appear at roughly 200 a day. The report says Qwen has become "part of the default workflow for developers deciding what models to fine-tune and deploy." Alibaba is also pushing Qwen through its cloud into Southeast Asia and Africa, markets where American labs have almost no presence, and where a very large share of the next generation of developers will learn to build. Meta and Nvidia have both rushed out new open models in recent weeks. That is what a response looks like when you feel the floor move. And to be clear, these are download and derivative numbers, not usage numbers. ChatGPT and Claude cannot be downloaded at all, so they do not appear in this comparison. What the figures measure is what developers choose to build on top of, which is a different question from what consumers type into a box. That is also why it matters MORE. Consumer habits change in an afternoon. Infrastructure choices last a decade, because everything built on top has to be rewritten to undo them. The American labs are valued on an assumption that frontier intelligence stays scarce, expensive, and rented by the token. Alibaba just made a version of it free, permanent, and small enough to run on hardware people already own. You will not get an announcement when the software you use every day starts running on a Chinese model underneath. Go and count how many of the tools you rely on could be rebuilt on free weights this year.

Ricardo

81,295 次观看 • 29 天前

A group of young women put a dog collar on a man and humiliate him. If you want to watch all the thousands of amateur femdom videos, send me a DM to join the VIP group 📩

Sensitive content

A group of young women put a dog collar on a man and humiliate him. If you want to watch all the thousands of amateur femdom videos, send me a DM to join the VIP group 📩

Cuckold Group

27,301 次观看 • 19 天前

Australian engineers have developed a unique virtual fencing system using a tactile collar for animal control. The farmer sets the boundaries of the virtual pasture through a smartphone app and the collar uses pulses and sound signals to guide the animals to the right place, eliminating the need for expensive electric fencing infrastructure.

Australian engineers have developed a unique virtual fencing system using a tactile collar for animal control. The farmer sets the boundaries of the virtual pasture through a smartphone app and the collar uses pulses and sound signals to guide the animals to the right place, eliminating the need for expensive electric fencing infrastructure.

Interesting STEM

164,138 次观看 • 8 个月前

Join us in NYC on June 10-11 for two days of hands-on learning, networking and inspiration. Connect with thousands of customers, partners, developers and tech leaders, all coming together for the observability and security event of the year. Register now:

Join us in NYC on June 10-11 for two days of hands-on learning, networking and inspiration. Connect with thousands of customers, partners, developers and tech leaders, all coming together for the observability and security event of the year. Register now:

Datadog, Inc.

19,519 次观看 • 1 年前

A little update video on one of last weeks rescues. The dog with an embedded collar. He's recovering well in the garden at Ernestos Sanctuary for Cats in Syria. Please keep him in your thoughts and prayers.

A little update video on one of last weeks rescues. The dog with an embedded collar. He's recovering well in the garden at Ernestos Sanctuary for Cats in Syria. Please keep him in your thoughts and prayers.

Ernesto’s Sanctuary for Syrian Cats

22,387 次观看 • 1 年前

Pi Network's new App Studio change and what it means Pi Network (Pi Network) made a major change to its App Studio pricing scheme. As expected, the change has become a huge talking point within the Pi ecosystem, with pioneers sharing their thoughts... However, the change means App Studio has moved from an experimental phase to one that aims to reward Pioneers for developing truly useful apps. Here are the key details of what exactly changed, why it happened, and what difference it makes: What Changed? Prior to yesterday's update, app creators paid a flat rate of 0.25 $PI for app creation using App Studio and another 0.25 PI to make any edits. This price was well below the underlying generative AI services the network uses. The new pricing, however, brought a significant change: 1.) The fees will be more reflective of the underlying computational and AI costs for creating and editing apps. 2.) The pricing will be dynamic... This means that more complicated apps and edits will be charged at higher costs. 3.) Pi Network will not mark up the underlying AI cost. 4.) Developers who still have some usage of their apps from unique users will continue to receive the 0.25 fee. In some cases, the new model cost reached 30.30 PI for app creation and 2.5 PI for editing, depending on the application type. Why did Pi Network make this change: 1.) Avoid wastage of resources. 2.) Focus on helping applications that provide real value and are being utilized by real pioneers. 3.) Provide creators with an economic reason to improve the quality and distribution of their applications. Simply put, the network has started selecting the serious creators. Cheap creation has resulted in a directory of apps that barely anyone opens. What the new model means for creators: The transition brings both positive and negative changes. 1.) Developers who already have some level of user activity will keep the old pricing model and retain their competitive edge. 2.) Other developers will have to pay higher prices. Early Pioneers who developed and polished their apps within the time frame had an opportunity to improve usability and attract users. Now developers must meet a certain threshold before they can use the platform. The price is no longer close to zero.

Pi Network's new App Studio change and what it means Pi Network (Pi Network) made a major change to its App Studio pricing scheme. As expected, the change has become a huge talking point within the Pi ecosystem, with pioneers sharing their thoughts... However, the change means App Studio has moved from an experimental phase to one that aims to reward Pioneers for developing truly useful apps. Here are the key details of what exactly changed, why it happened, and what difference it makes: What Changed? Prior to yesterday's update, app creators paid a flat rate of 0.25 $PI for app creation using App Studio and another 0.25 PI to make any edits. This price was well below the underlying generative AI services the network uses. The new pricing, however, brought a significant change: 1.) The fees will be more reflective of the underlying computational and AI costs for creating and editing apps. 2.) The pricing will be dynamic... This means that more complicated apps and edits will be charged at higher costs. 3.) Pi Network will not mark up the underlying AI cost. 4.) Developers who still have some usage of their apps from unique users will continue to receive the 0.25 fee. In some cases, the new model cost reached 30.30 PI for app creation and 2.5 PI for editing, depending on the application type. Why did Pi Network make this change: 1.) Avoid wastage of resources. 2.) Focus on helping applications that provide real value and are being utilized by real pioneers. 3.) Provide creators with an economic reason to improve the quality and distribution of their applications. Simply put, the network has started selecting the serious creators. Cheap creation has resulted in a directory of apps that barely anyone opens. What the new model means for creators: The transition brings both positive and negative changes. 1.) Developers who already have some level of user activity will keep the old pricing model and retain their competitive edge. 2.) Other developers will have to pay higher prices. Early Pioneers who developed and polished their apps within the time frame had an opportunity to improve usability and attract users. Now developers must meet a certain threshold before they can use the platform. The price is no longer close to zero.

BSCN

41,305 次观看 • 20 天前

Today we're introducing TRIBE v2 (Trimodal Brain Encoder), a foundation model trained to predict how the human brain responds to almost any sight or sound. Building on our Algonauts 2025 award-winning architecture, TRIBE v2 draws on 500+ hours of fMRI recordings from 700+ people to create a digital twin of neural activity and enable zero-shot predictions for new subjects, languages, and tasks. Try the demo and learn more here:

Today we're introducing TRIBE v2 (Trimodal Brain Encoder), a foundation model trained to predict how the human brain responds to almost any sight or sound. Building on our Algonauts 2025 award-winning architecture, TRIBE v2 draws on 500+ hours of fMRI recordings from 700+ people to create a digital twin of neural activity and enable zero-shot predictions for new subjects, languages, and tasks. Try the demo and learn more here:

AI at Meta

6,975,687 次观看 • 5 个月前

Sam Altman announces the OpenAI Academy, a program to give developers and organizations in low- and middle-income countries access to the latest AI tools, as part of the Partnership for Global Inclusivity on AI

Sam Altman announces the OpenAI Academy, a program to give developers and organizations in low- and middle-income countries access to the latest AI tools, as part of the Partnership for Global Inclusivity on AI

Tsarathustra

313,034 次观看 • 2 年前

This is the best and fastest speech-to-text model in the world: • 23.2 seconds to process 30 minutes of audio • 93.3% accuracy • Diarization support to detect multiple speakers • Trained on 12.5 million hours of multilingual data I tried it out and it's pretty impressive:

This is the best and fastest speech-to-text model in the world: • 23.2 seconds to process 30 minutes of audio • 93.3% accuracy • Diarization support to detect multiple speakers • Trained on 12.5 million hours of multilingual data I tried it out and it's pretty impressive:

Santiago

66,420 次观看 • 1 年前

Action Model went live a few hours ago, and thousands of you have already signed up to help build an AI model owned by the people. If you haven’t joined yet, now’s the time. The future of AI isn’t corporate-owned. It’s community-owned. Here’s a quick tutorial on exactly how to sign up for Action Model 👇

Action Model went live a few hours ago, and thousands of you have already signed up to help build an AI model owned by the people. If you haven’t joined yet, now’s the time. The future of AI isn’t corporate-owned. It’s community-owned. Here’s a quick tutorial on exactly how to sign up for Action Model 👇

Georgia | Action Model

22,071 次观看 • 7 个月前

13. That man deserves heaven on earth.💕 The happiness of the dogs when they saw the man who set up a shelter and food for paralyzed dogs. 🐕

13. That man deserves heaven on earth.💕 The happiness of the dogs when they saw the man who set up a shelter and food for paralyzed dogs. 🐕

Awesome Videos ❤️

220,029 次观看 • 1 年前

Introducing Parrot 🦜 Our first speech-to-text model built for production-grade voice agents with an industry leading accuracy, built for noisy calls, regional dialects and compressed audio. Trained on thousands of hours of real-world audio to make sure when you say Udaipur, the agent doesn't hear 'you-die-poor'.

Introducing Parrot 🦜 Our first speech-to-text model built for production-grade voice agents with an industry leading accuracy, built for noisy calls, regional dialects and compressed audio. Trained on thousands of hours of real-world audio to make sure when you say Udaipur, the agent doesn't hear 'you-die-poor'.

Vedant

15,953 次观看 • 3 个月前