Meta joins AI chatbot race with own large language model for researchers

by Hansraj Agrawal | Feb 25, 2023 | Technology

After Microsoft ChatGPT and Google’s Bard, Meta is joining the AI chatbot race with its own state-of-the-art foundational large language model designed to help researchers advance their work in the field of artificial intelligence.

However, Meta’s Large Language Model Meta AI (LLaMA) isn’t like ChatGPT-driven Bing at the moment as it can’t yet talk to humans but will help researchers.

“Smaller, more performant models such as LLaMA enable others in the research community who don’t have access to large amounts of infrastructure to study these models, further democratising access in this important, fast-changing field,” Meta said in a statement.

Meta is making LLaMA available at several sizes (7 billion, 13 billion, 33 billion, and 65 billion parameters).

Large language models — natural language processing (NLP) systems with billions of parameters — have shown new capabilities to generate creative text, solve mathematical theorems, predict protein structures, answer reading comprehension questions, and more.

“They are one of the clearest cases of the substantial potential benefits AI can offer at scale to billions of people,” said Meta.

Smaller models trained on more tokens — which are pieces of words — are easier to retrain and fine-tune for specific potential product use cases.

Meta has trained LLaMA 65 billion and LLaMA 33 billion on 1.4 trillion tokens.

“Our smallest model, LLaMA 7B, is trained on one trillion tokens,” said the company.

Like other large language models, LLaMA works by taking a sequence of words as an input and predicts a next word to recursively generate text.

“To train our model, we chose text from the 20 languages with the most speakers, focusing on those with Latin and Cyrillic alphabets,” Meta informed.

To maintain integrity and prevent misuse, the company said it is releasing the model under a noncommercial license focused on research use cases at the moment.

–IANS

na/ksk/

(Only the headline and picture of this report may have been reworked by the Business Standard staff; the rest of the content is auto-generated from a syndicated feed.)

Source link

ChatGPT-driven smart home voice assistant coming soon: Josh.ai co-founder

by Hansraj Agrawal | Feb 25, 2023 | Technology

US-based artificial intelligence company Josh.ai, which is known for developing the voice-controlled home automation system, has started working on a prototype integration using OpenAI’s ChatGPT.

Turn on the lights, how is the temperature, you might have asked such questions to your voice assistants like Alexa or Siri, but instead of such questions, imagine your voice assistant could also respond to nebulous comments like “I’ve had a tough day; What’s a good way to relax?

According to Alex Capecelatro, co-founder of the Josh.ai home automation system, that’s the potential of voice assistants powered by new AI language models.

“We are thrilled to be working on bringing the best of Josh.ai and ChatGPT together to create something truly remarkable – a solution where one plus one equals three. By combining our strengths, we envision delivering an AI experience that is beyond what any smart home is capable of,” he said.

Moreover, Capecelatro explained by giving some examples of how ChatGPT-enabled voice assistant would work.

“Ok Josh, tell me a bedtime story”, where Josh.ai + ChatGPT will provide stories based on the location of the home and other factors unique to the family.

“Ok Josh, the kids are coming in and it’s getting dark can you make sure the kitchen is ready for them?” where Josh.ai + ChatGPT can properly prepare the space.

–IANS

shs/svn/

(Only the headline and picture of this report may have been reworked by the Business Standard staff; the rest of the content is auto-generated from a syndicated feed.)

Source link

Microsoft conducts research on use of ChatGPT to instruct robots, drones

by Hansraj Agrawal | Feb 24, 2023 | Technology

In order to test if people can use ChatGPT to instruct robots without knowing programming languages or understanding robotics systems, Microsoft has conducted research to see if it can think outside the text and think about the physical world to help with robotics tasks.

“The key challenge here is teaching ChatGPT how to solve problems considering the laws of physics, the context of the operating environment, and how the robot’s physical actions can change the state of the world,” Microsoft said in a blogpost.

“It turns out that ChatGPT can do a lot by itself, but it still needs some help. Our technical paper describes a series of design principles that can be used to guide language models towards solving robotics tasks. These include, and are not limited to, special prompting structures, high-level APIs, and human feedback via text,” it added.

After ChatGPT was given access to object-detection and object-distance data via application interfaces, the Microsoft researchers investigated its ability to generate code, mostly in Python, for robotics scenarios such as zero-shot planning and code generation.

As the AI chatbot was trained on a large amount of code and written text, it can generate code.

The system can solve coding problems and debug programmes, as well as respond to dialogue and seek clarifications, according to the researchers.

Moreover, ChatGPT was tested as a language-based interface between a non-technical user and a drone using these dialogue and clarification capabilities in mind.

The researchers note that while GPT-3, LaMDA, and Codex demonstrated good results when it comes to robotics planning and coding, ChatGPT stands out as “a potentially more versatile robotics tool because it incorporates natural language and code generation models with dialogue flexibility”.

“ChatGPT asked clarification questions when the user’s instructions were ambiguous and wrote complex code structures for the drone such as a zig-zag pattern to visually inspect shelves,” the researchers said.

Microsoft tested ChatGPT to use a robotic arm to move blocks around to form the Microsoft logo, with writing an algorithm for a drone to reach a point without crashing into obstacles, and whether ChatGPT can decide where a robot should go based on sensor feedback in real-time.

–IANS

shs/vd

(Only the headline and picture of this report may have been reworked by the Business Standard staff; the rest of the content is auto-generated from a syndicated feed.)

Source link

Amazon’s Alexa introduces new male voice as it completes 5 years in India

by Hansraj Agrawal | Feb 24, 2023 | Technology

Amazon’s Alexa is completing 5 years in India, and to celebrate the occasion Amazon has introduced new male voice options for Alexa’s responses. Indian users can now switch between the original and a new male voice. Also, Alexa can now speak in both Hindi and English.

Dilip R S, Amazon’s country manager for Alexa, revealed in a statement, “Over the past five years, our aim has been to build Alexa from India and for India, and I truly believe that our journey is synonymous with the evolution of ambient computing in the country. As we look ahead, our focus remains on bringing to life newer voice, touch, motion, and vison-enabled experiences for consuming entertainment, completing tasks, and accessing information.”

How to change Voice in Alexa?

Alexa’s dual voice has brought excitement to Indian users. If you have an Alexa device at your home and haven’t figured out how to change Alexa’s voice, here’s how you can do it:

The company has mentioned that users can change Alexa’s voice by saying, “Alexa, change your voice” on the echo device.

Or,

You can also change the voice via the Alexa app by going to Alexa’s individual device settings and selecting the voice of your choice.

Amazon Alexa’s gain in popularity

In a very short period, Alexa has gained massive popularity in India. The company revealed last year that requests to Alexa for music went up by 53 per cent as customers streamed their favourite songs through different platforms such as Amazon Prime Music, Spotify, Jio Savan, and Apple Music.

As people want more comfort, the request to control home appliances has increased tremendously.

The company also said that users loving to interact with Alexa was reflected in their questions, such as “Alexa, how are you” (31,680 times/day), and “Alexa, I love you” (21,600 times/day).

Source link

Qualcomm demonstrates world’s first stable diffusion on Android phone

by Hansraj Agrawal | Feb 24, 2023 | Technology

Chip maker Qualcomm has showcased the world’s first demonstration of Stable Diffusion on an Android device to further improve its efforts in working with AI.

Stable Diffusion is a deep learning, text-to-image foundation model, first created in 2022 and is commonly used to generate detailed images based on text descriptions within tens of seconds, as seen on ‘DALL-E’ and other similar online platforms.

For the first time, Qualcomm’s AI Research team performed full-stack AI optimisations using the company’s AI Stack to deploy Stable Diffusion on an Android smartphone powered by its Snapdragon 8 Gen 2 platform.

“For Stable Diffusion, we started with the FP32 version 1-5 open-source model from Hugging Face and made optimisations through quantisation, compilation, and hardware acceleration to run it on a phone powered by Snapdragon 8 Gen 2 Mobile Platform,” the company said in a blogpost.

Qualcomm’s AI stack software employs artificial intelligence with support for a diverse range of intelligent devices, including mobile, automotive, IoT, cloud platforms, and others.

This demonstration essentially allowed a smartphone to generate an image render based on text descriptions in less than 15 seconds.

“Our one technology roadmap allows us to scale and utilise a single AI stack that works across not only different end devices but also different models,” Qualcomm said.

“This means that the optimisations for Stable Diffusion to run efficiently on phones can also be used for other platforms like laptops, XR headsets, and virtually any other device powered by Qualcomm Technologies,” it added.

–IANS

shs/uk/

(Only the headline and picture of this report may have been reworked by the Business Standard staff; the rest of the content is auto-generated from a syndicated feed.)

Source link

Video-sharing platform YouTube to allow dubbing videos in multi-languages

by Hansraj Agrawal | Feb 24, 2023 | Technology

Video-sharing platform YouTube has announced that it is rolling out support for multi-language audio tracks, which will allow creators to dub their new and existing videos in different languages.

Over the past year, the company was testing this feature with popular YouTube content creator Jimmy Donaldson aka MrBeast and a small group of creators.

“For viewers, multi-language audio means they can now watch videos dubbed in their primary language, introducing them to even more content that they otherwise may not have seen,” the company said in a blogpost on Thursday.

“We found that creators testing multi-language dubbed videos saw over 15 per cent of their watch time coming from views in the video’s non-primary language,” it added.

Moreover, on average, viewers watched over two million hours of dubbed video daily in the last month alone.

More than 3,500 multi-language videos are already uploaded in over 40 languages, the platform said.

Creators have to add different audio tracks through the “Subtitles Editor” tool when uploading a video to their channel.

The company further mentioned that “existing content in creators’ catalogues can be updated with additional audio tracks as well.”

“If you’re a viewer, just click the video’s settings to see what audio tracks are available to start watching in another language,” it added.

(Only the headline and picture of this report may have been reworked by the Business Standard staff; the rest of the content is auto-generated from a syndicated feed.)

Source link

Times Top News

Times Top News