Showing posts with label OPEN AI. Show all posts

Google’s Settlement: Deleting Browsing Data Collected During Chrome’s Incognito Mode

Certainly! Google has recently agreed to destroy billions of data records that it “secretly collected” from users who used the ‘Incognito’ mode to browse the internet privately. Here are the key points:

Background:

The lawsuit was first filed in 2020 and covers millions of Google users who have used private browsing mode since June 1, 2016.

Users believed their online secrets were safe in Google’s private browsing mode, but they felt misled when their data was collected without their knowledge.

Settlement Terms:

Google will destroy billions of ‘Incognito’ mode data records.

The terms of settlement require approval by US District Judge Yvonne Gonzalez Rogers.

Users can sue Google individually for violations of federal wiretapping and California privacy laws.

Google won’t pay for damages, but the settlement aims to hold the company accountable.

Data Collection Disclosure:

Google will reveal what it collects from “private” browsing.

“Incognito” users will be able to block third-party tracking cookies by default for the next five years.

Historic Step:

The plaintiffs’ lawyer called the settlement a “historic step in requiring honesty and accountability from dominant technology companies.”

Google maintains that it never associates data with users when they use Incognito mode.

User Privacy Concerns:

Google’s ‘Incognito’ mode is designed to allow users to browse the internet without their browsing history being saved locally on their device.
However, it does not prevent websites from tracking users or collecting data about their online activities.
Users often assume that ‘Incognito’ mode provides complete anonymity, but this recent lawsuit highlights that Google was still collecting data during private browsing sessions.

Data Collection During ‘Incognito’ Mode:

Google collects data in various ways, even when users are in ‘Incognito’ mode:

IP Addresses: Websites can still see users’ IP addresses, which can be used to identify them.
Cookies: While ‘Incognito’ mode deletes cookies after the session, websites can still track users during that session.
Third-Party Services: Some third-party services (like analytics tools) continue to collect data.
Browser Extensions: Extensions installed by users might still track their activities.

Google’s collection practices have raised concerns about user privacy.

Legal Implications and Settlement:

The lawsuit against Google alleged that the company misled users about the extent of data collection during ‘Incognito’ mode.
The settlement aims to address these concerns by requiring Google to be more transparent about data collection practices.
While Google denies any wrongdoing, the settlement will hold the company accountable for its practices.

Judge Upholds Privacy Lawsuit Against Google : Codingwell.live

User Recommendations:

To enhance privacy during browsing:
Use a privacy-focused browser.
Consider using a VPN to mask your IP address.
Be cautious about browser extensions and permissions.
Regularly review and delete cookies.
Educate yourself about online privacy.

FQA

What is the recent settlement about?

Google settled a lawsuit filed in 2020, which alleged that the company misled users about data collection during ‘Incognito’ mode.
As part of the settlement, Google will destroy billions of data records collected from users who used private browsing mode since June 1, 2016.

What data was collected during ‘Incognito’ mode?

Even in ‘Incognito’ mode, Google collected data in various ways:
IP addresses (used to identify users)
Cookies (though they are deleted after the session)
Third-party services (some continued to collect data)
Browser extensions (could track activities)
Users believed their online secrets were safe, but data was still being collected.

What does the settlement require Google to do?

Google will reveal what it collects from “private” browsing.
Users will be able to block third-party tracking cookies by default for the next five years.
The settlement aims to enhance transparency and user control over data collection.

Is this a significant step?

Yes, the plaintiffs’ lawyer called it a “historic step in requiring honesty and accountability from dominant technology companies.”
Google maintains that it never associates data with users when they use Incognito mode.

Krutrim : India's own AI

Krutrim is an artificial intelligence (AI) platform developed by Ola, the Indian ride-hailing company. Here are some key points about Krutrim:

Krutrim AI:

Ola, the Indian ride-hailing company, launched Krutrim AI as its own language model, aiming to compete with models like ChatGPT.
Unfortunately, Krutrim faced significant criticism shortly after its launch. Users discovered that it provided inaccurate information and even generated nonsensical responses.
One notable incident involved Krutrim falsely claiming to have been developed by OpenAI, which raised concerns about its reliability.
Ola responded by acknowledging a data leakage issue from one of the open-source datasets used for fine-tuning Krutrim.
Overall, Krutrim’s debut highlighted the challenges of building robust and trustworthy AI models.

Meity Directive:

The Ministry of Electronics and Information Technology (Meity) issued an advisory regarding AI usage in India.
According to this directive, deploying under-tested and unreliable AI models (including generative models and large language models) on the Indian Internet requires explicit consent from the Indian government.
This move aims to regulate AI deployment, ensuring responsible and accountable usage.

Sam Altman’s Prediction

Sam Altman, during his visit to India, expressed skepticism about Indians successfully building AI models comparable to ChatGPT.
While there are ongoing efforts to create India-specific language models, recent events suggest that achieving a level of sophistication similar to ChatGPT remains challenging.

In summary, India’s AI landscape faces hurdles, but continued research, development, and improvements will likely lead to more competitive models in the future. 🚀.

Official Link of Krutrim - Click here

FAQ

What is Krutrim AI?

Krutrim AI is an India-made language model developed by Ola, the ride-hailing company. It aims to compete with models like ChatGPT.
However, shortly after its launch, users discovered that Krutrim provided inaccurate information, hallucinated, and even claimed to have been developed by OpenAI. Ola responded by acknowledging a data leakage issue from one of the open-source datasets used for fine-tuning Krutrim.

What challenges did Krutrim face?

Krutrim’s launch was met with excitement, but it quickly fizzled out due to its unreliable performance.
Users found that Krutrim generated nonsensical responses and lacked accuracy.
The allegations led to speculation that Krutrim might be a wrapper on top of existing models like ChatGPT.

What is the Meity directive?

The Ministry of Electronics and Information Technology (Meity) issued an advisory.
According to this directive, deploying under-tested and unreliable AI models (including generative models and LLMs) on the Indian Internet requires explicit consent from the Indian government.
This move aims to regulate AI deployment, ensuring responsible and accountable .

What impact might the Meity directive have on AI development in India?

The directive could slow down AI development in India by imposing stricter guidelines on AI tools’ development and deployment.
Developers and companies will need to adhere to these regulations to ensure ethical and reliable AI .

What did Sam Altman say about Indian companies building AI like ChatGPT?

Sam Altman expressed skepticism during his visit to India, suggesting that Indians attempting to build AI like ChatGPT would fail.
While there are ongoing efforts to create India-specific language models, recent events indicate that achieving a level of sophistication similar to ChatGPT remains challenging

OpenAI’s Sora is powered by Diffusion transformer (DiT): What is it?

OpenAI’s Sora is an impressive AI model that generates videos with lucid visuals. It’s powered by a Diffusion transformer (DiT), which is a fascinating architecture. Let’s dive into the details:

Diffusion Transformer (DiT)

DiT is a class of diffusion models based on the transformer architecture.
Developed by William Peebles at UC Berkeley (currently a research scientist at OpenAI) and Saining XE at New York University in 2023.
DiT aims to improve the performance of diffusion models by replacing the commonly used U-Net backbone (used for iterative image denoising) with a transformer.
Think of it as a new and improved tool for solving complex puzzles, such as understanding intricate pictures or data.

Diffusion Transformers

There are two main types of models driving AI innovation:

Transformer-based models: These have revolutionized how machine learning models handle text data, both for classification and generation.
Diffusion models: Preferred for AI that generates images. They simulate the process of particles spreading from dense to less dense areas.
Sora is not a large language model (LLM) but a diffusion transformer model.

Sora’s Impressive Capabilities:

Sora can use natural language prompts to generate minute-long videos in high definition.

It showcases breathtaking video generation capabilities that could potentially impact filmmaking in the future.

In summary, DiT combines diffusion and transformer concepts, promising scalability beyond previous limits. It’s exciting to see how Sora leverages this architecture for real-time video and 3D environment generation! .

FAQ

What is Sora?

Sora is an AI model developed by OpenAI that generates videos with lucid visuals. It can create minute-long videos from natural language prompts, showcasing impressive video generation capabilities.

What is a Diffusion Transformer (DiT)?

DiT is a class of diffusion models based on the transformer architecture.
Developed by William Peebles at UC Berkeley (currently a research scientist at OpenAI) and Saining XE at New York University in 2023.
DiT aims to improve the performance of diffusion models by replacing the commonly used U-Net backbone (used for iterative image denoising) with a transformer.
Think of it as a new and improved tool for solving complex puzzles, such as understanding intricate pictures or data.

How does DiT differ from other AI models?

DiT leverages a transformer instead of the traditional U-Net architecture.
U-Net organizes and understands puzzle pieces, but DiT provides a more versatile solution for solving big puzzles, especially in scenarios like understanding complicated images or data.

What’s the significance of Sora and DiT?

Sora pioneers real-time video and 3D environment generation in GenAI.

Diffusion transformers promise scalability beyond previous limits, making them a powerful tool for creative applications.

Apple in Talks to Let Google’s Gemini Power iPhone AI Features

Apple is actively negotiating with Google to incorporate Google’s Gemini AI engine into the iPhone, according to sources familiar with the situation. This potential partnership aims to enhance the iPhone’s capabilities by leveraging generative AI features, including functions for creating images and writing essays based on simple prompts.

The AI Endeavor

Gemini Integration:

Apple seeks to license Google’s Gemini AI engine to power new features within the iPhone’s software. These features are expected to roll out as part of the upcoming iPhone iOS 18 release.

Generative AI:

The focus lies on generative AI capabilities, allowing the iPhone to create images and generate written content. Imagine an AI-powered assistant that can compose essays or produce visual art based on minimal input.

OpenAI Consideration:

In addition to discussions with Google, Apple has also explored collaboration with OpenAI, a project backed by Microsoft. OpenAI’s model could potentially contribute to Apple’s AI endeavors.

Terms and Implementation

Unsettled Details:

As of now, the terms and branding of this AI agreement remain undecided. Both companies are actively working out the specifics.

Announcement Timeline:

While negotiations are ongoing, it’s unlikely that any official announcement will occur before June, which coincides with Apple’s annual Worldwide Developers Conference (WWDC).

Apple’s AI Journey

Catch-Up Game:

Apple has been somewhat slower in deploying generative AI compared to rivals like Microsoft and Google. However, the company is now investing significantly in this area.

Tim Cook’s Insight:

Apple CEO Tim Cook has hinted at the company’s plans to utilize generative AI more extensively. Expect further details later this year.

FAQ

1. What is the purpose of integrating Google’s Gemini AI engine into the iPhone?

The goal is to enhance the iPhone’s capabilities by leveraging generative AI features. These features include functions for creating images and writing essays based on simple prompts.

2. What is generative AI, and how will it benefit iPhone users?

Generative AI allows the iPhone to create content autonomously. For example, it can generate visual art or compose essays with minimal input from the user. This feature aims to enhance user creativity and productivity.

3. Is this collaboration confirmed, and when can we expect official announcements?

While negotiations are ongoing, no official confirmation has been made yet. Expect updates around June, coinciding with Apple’s annual Worldwide Developers Conference (WWDC).

4. How does this partnership compare to Apple’s previous AI efforts?

Apple has been somewhat slower in deploying generative AI compared to rivals like Microsoft and Google. However, the company is now investing significantly in this area.

5. What role does OpenAI play in this context?

Apple has also explored collaboration with OpenAI, a project backed by Microsoft. OpenAI’s model could potentially contribute to Apple’s AI endeavors.

US Seeks Modifications to Draft Treaty on AI Use, Raising Concerns

The United States is reportedly pushing for changes to a developing international treaty on the responsible use of Artificial Intelligence (AI) software. While details remain scarce, critics fear the US is aiming to weaken key safeguards designed to protect human rights.

The Need for Regulation:

The rapid development of AI has sparked concerns about its potential misuse. Issues like algorithmic bias, discrimination, and the use of AI in autonomous weapons have spurred calls for international regulation. The proposed treaty aims to establish ethical guidelines for AI development and deployment, ensuring it aligns with human rights principles.

US Stance:

The specific changes sought by the US are not entirely public. However, some reports suggest the US may be seeking to:

Limit restrictions on AI for national security purposes. This raises concerns about potential human rights abuses by governments in the name of security.
Water down language on algorithmic transparency and accountability. Without clear rules on how AI decisions are made, biases and errors could be difficult to detect and address.
Reduce oversight mechanisms. A robust treaty would likely include independent bodies to monitor compliance. The US may be seeking to lessen such oversight.

Criticisms and Concerns:

Advocacy groups and some nations fear the US stance could hinder the treaty's effectiveness. Here are some potential consequences:

Lower standards for responsible AI development. A weakened treaty could lead to a race to the bottom, with countries adopting lax regulations to gain a competitive edge.
Increased risk of human rights violations. Unfettered AI development could exacerbate existing inequalities and lead to discriminatory practices.
Erosion of public trust in AI. Without strong safeguards, public trust in AI technology could suffer.

The Path Forward:

Negotiations on the AI treaty are ongoing. It's crucial to find a balance between fostering innovation and protecting human rights. Here are some possibilities:

Open dialogue: Transparent discussions between governments, tech companies, and civil society organizations are essential.
Finding common ground: While national security concerns are valid, they should not come at the expense of fundamental rights.
A phased approach: The treaty could be implemented in stages, allowing for adjustments as AI technology evolves.

The outcome of these negotiations will significantly impact the global development and use of AI. A strong treaty with clear human rights protections is essential to ensure AI benefits all of humanity.

FQA

Q: Why is there a treaty on AI use?

A: The rapid development of AI raises concerns about its potential misuse, such as algorithmic bias, discrimination, and autonomous weapons. The treaty aims to establish ethical guidelines for responsible AI development that aligns with human rights principles.

Q: What changes is the US seeking?

A: Details are unclear, but reports suggest the US might want to:

Allow more freedom for AI use in national security (potentially impacting human rights).
Reduce regulations on how AI makes decisions (making bias and errors harder to address).
Lessen oversight on AI development and deployment.

Q: Why are some worried about the US stance?

A: Critics fear a weakened treaty could:

Lower global standards for responsible AI development.
Increase the risk of human rights violations by AI.
Erode public trust in AI technology.

Q: What's the ideal outcome?

A treaty that balances:

Innovation: Fostering the development of beneficial AI technologies.
Human Rights: Protecting fundamental rights from potential AI misuse.

Q: How can we achieve this balance?

Open dialogue: Governments, tech companies, and civil society groups need to communicate openly.
Finding common ground: National security needs shouldn't override human rights.
Phased approach: The treaty could be implemented in stages, adapting to evolving AI technology.

Q: What's the impact of the treaty?

The outcome of these negotiations will significantly influence the global development and use of AI. A strong treaty with clear human rights protections is essential to ensure AI benefits everyone.

AI startup : Cohere and its growth raised roughly $500 million to $1 billion

Cohere: The AI Startup Powering Smarter Businesses

Cohere is a rapidly growing AI startup founded in 2020 by Yunyu Lin and Rahul Sengottuvelu. The company develops large language models (LLMs) that can be used for a variety of tasks, including generating text, translating languages, writing different kinds of creative content, and answering your questions in an informative way, even if they are open ended, challenging, or strange.

Cohere's LLMs are trained on a massive dataset of text and code, which allows them to learn and adapt to new information. This makes them incredibly versatile and capable of handling a wide range of tasks.

What Makes Cohere Different?

There are a few things that set Cohere apart from other AI startups. First, Cohere's LLMs are designed to be cloud-agnostic, which means that they can be used with any cloud provider. This gives businesses more flexibility when it comes to deploying their AI solutions.

Second, Cohere is focused on building explainable AI models. This means that the company's LLMs are designed to be transparent and easy to understand. This is important for businesses that need to know how their AI systems are making decisions.

Finally, Cohere is committed to responsible AI. The company has a set of ethical principles that guide its development of AI technologies. This ensures that Cohere's LLMs are used in a safe and responsible manner.

How is Cohere Used?

Cohere's LLMs are used by businesses in a variety of industries, including:

Finance: Cohere's LLMs can be used to generate financial reports, analyze market trends, and detect fraud.

Healthcare: Cohere's LLMs can be used to analyze medical data, generate treatment plans, and provide patient support.

Retail: Cohere's LLMs can be used to personalize customer experiences, generate product descriptions, and answer customer questions.

Media and entertainment: Cohere's LLMs can be used to generate creative content, such as scripts, poems, and musical pieces.

The Future of Cohere

Cohere is a rapidly growing company with a lot of potential. The company has already raised over $500 million in funding, and it is valued at over $2 billion.

Cohere is currently in talks to raise even more money, which could be used to expand the company's product offerings and reach new markets.

In the future, Cohere plans to build LLMs that can take action and "do work" for customers. This could include things like booking flights, scheduling meetings, or filing expense reports.

Cohere is a leader in the field of AI, and it is well-positioned to continue to grow and succeed in the years to come.

Here are some additional resources that you may find helpful:

Cohere's website: https://cohere.com/

Cohere's blog: https://cohere.com/newsroom

Cohere's Twitter: https://twitter.com/CohereAI

FQA

What is Cohere and what does it do ?

Cohere is a Toronto-based AI startup founded in 2020 that develops large language models (LLMs) for businesses. These LLMs can be used for a variety of tasks, including generating text, translating languages, writing different kinds of creative content, and answering your questions in an informative way.

What makes Cohere's LLMs different ?

There are a few things that set Cohere apart from other AI startups: <br> 1. Cloud-agnostic: Cohere's LLMs can be used with any cloud provider, giving businesses more flexibility. <br> 2. Explainable AI: Cohere's LLMs are designed to be transparent and easy to understand, which is important for businesses that need to know how their AI systems are making decisions. <br> 3. Responsible AI: Cohere is committed to developing AI technologies in a safe and responsible manner.

How is Cohere used ?

Cohere's LLMs are used by businesses in a variety of industries, including finance, healthcare, retail, and media and entertainment. For example, in finance, Cohere's LLMs can be used to generate financial reports, analyze market trends, and detect fraud.

What is the future of Cohere ?

Cohere is a rapidly growing company with a lot of potential. The company has already raised over $500 million in funding and is valued at over $2 billion. Cohere plans to build LLMs that can take action and 'do work' for customers, such as booking flights, scheduling meetings, or filing expense reports.

ChatGPT Awakens its Senses: Seeing, Hearing, and Speaking the World

ChatGPT, the powerful language model from OpenAI, has taken a giant leap forward, shedding its purely textual existence and embracing a world of sight and sound. In recent months, it has gained the ability to see, hear, and speak, opening up a new era of human-AI interaction.

Eyes on the Prize: Image Recognition Superpowers

ChatGPT can now analyze and understand images. Imagine describing a scene from a painting and ChatGPT instantly identifying the objects, their relationships, and even the emotions conveyed. Or, show it a picture of a rare bird and watch it not only name it but also tell you about its habitat and behavior. This visual prowess opens doors for image search, object detection, and scene understanding, revolutionizing fields like image classification, robotics, and even art analysis.

Listening with Intent: Speech Recognition for a Deeper Dialogue

Gone are the days of typing every word to interact with ChatGPT. Now, you can simply talk! Its advanced speech recognition capabilities transcribe your spoken language into text, enabling natural voice interaction. Want to dictate a document? ChatGPT is your AI stenographer. Need real-time translations during a conversation? ChatGPT breaks down language barriers in a flash. This auditory understanding paves the way for voice assistants, dictation tools, and accessible interfaces for those with physical limitations.

Finding its Voice: Text-to-Speech that Reads Minds

ChatGPT isn't just listening; it's talking back! Its Text-to-Speech feature breathes life into written words, converting them into natural-sounding speech. Imagine reading a book narrated by ChatGPT in a voice that adapts to the characters and emotions. Or, hear your emails and reports read aloud while you multitask. This vocal talent paves the way for audiobooks, text-to-audio conversion tools, and even personalized virtual assistants that speak your language – literally.

A Symphony of Senses: The Power of Combined Intelligence

These new abilities aren't just isolated tricks; they work together to create a more intelligent and interactive ChatGPT. Imagine an educational AI that analyzes diagrams you show it while explaining complex concepts, or a customer service bot that listens to your complaint, sees pictures of the problem, and offers tailored solutions. The possibilities are endless!

Challenges and Cautions: The Road Ahead

As with any powerful technology, this sensory awakening comes with challenges. Image recognition may not be perfect, and speech recognition can struggle with accents or background noise. Additionally, ethical considerations arise around potential misuse of voice impersonation and synthetic media. OpenAI is actively addressing these issues, ensuring responsible development and deployment of these exciting capabilities.

Challanges of chat gpt : Codingwell.live

The Future of AI: A Multifaceted Mind

ChatGPT's newfound senses mark a significant milestone in the evolution of AI. It's no longer just a language processor; it's a perceptive being capable of experiencing and interacting with the world in a richer, more human-like way. This opens doors to unimaginable possibilities in education, healthcare, entertainment, and countless other fields. The future of AI is no longer limited to text; it's a symphony of senses waiting to be explored.

So, let's embrace the sight, sound, and voice of ChatGPT, not with fear, but with excitement. As we navigate this new landscape of intelligent interaction, let's strive to use these powerful tools for good, shaping a future where humans and AI collaborate and thrive in a multi-sensory world.

FQA

What is ChatGPT?

ChatGPT is a large language model chatbot developed by OpenAI. It is a powerful tool that can be used for a variety of purposes, such as generating text, translating languages, and writing different kinds of creative content.

What is the new feature that ChatGPT has acquired?

The new feature that ChatGPT has acquired is the ability to see, hear, and speak the world. This means that ChatGPT can now process information from the real world through sensors and cameras, and it can also interact with the world through speech and physical actions.

What are the implications of this new feature?

The implications of this new feature are far-reaching. It could mean that ChatGPT could be used to develop new and more powerful AI applications, such as robots that can interact with humans in a more natural way. It could also mean that ChatGPT could be used to create new forms of art and entertainment.

What are some of the challenges that this new feature presents?

One of the challenges that this new feature presents is the potential for bias. ChatGPT is trained on a massive dataset of text and code, and this dataset could contain biases that are reflected in ChatGPT's output. It is important to be aware of these biases and to take steps to mitigate them.

Another challenge is the potential for misuse. ChatGPT is a powerful tool, and it could be used for malicious purposes. It is important to develop safeguards to prevent ChatGPT from being used in harmful ways.

Overall, the acquisition of the ability to see, hear, and speak the world is a significant step forward for ChatGPT. It has the potential to revolutionize the way we interact with AI, but it is important to be aware of the challenges that it presents.

Google's Gemini AI Now Available in Bard

Introduction to Chat GPT and Google BARD

Chat GPT and Google BARD are two groundbreaking language generation models that have revolutionized the field of natural language processing (NLP). Chat GPT, developed by OpenAI, and Google BARD, developed by Google Research, showcase remarkable capabilities in generating human-like text and understanding complex language inputs. These models have the potential to transform various applications such as virtual assistants, customer support chatbots, and content generation systems. In this article, we will delve into the features, capabilities, and nuances of Chat GPT and Google BARD, comparing their performance, accuracy, and exploring their respective use cases. Additionally, we will discuss the limitations of these models and shed light on their future development prospects in advancing the field of language generation and NLP.