OTHER
Why Meta is Allowing Users to See the Inner Workings of Its New AI Chatbot
The AI division of Mark Zuckerberg’s Meta recently unveiled its Llama 2 chatbot. Microsoft has been appointed as Meta’s preferred partner on Llama 2, which will be available through the Windows operating system.
Meta’s approach with Llama 2 contrasts with that of the company OpenAI, which created the AI chatbot ChatGPT. That’s because Meta has made its product open source – meaning that the original code is freely available, allowing it to be researched and modified.
This strategy has sparked a vast wave of discussions. Will it foster greater public scrutiny and regulation of large language models (LLMs) – the technology that underlies AI chatbots such as Llama 2 and ChatGPT? Could it inadvertently empower criminals to use the technology to help them carry out phishing attacks or develop malware? And could the move help Meta gain an advantage over OpenAI and Google in this fast-moving field? Whatever happens, this strategic move looks set to reshape the current landscape of generative AI. In February 2023, Meta released its first version of the LLM, called Llama, but made it available for academic use only. Its updated version, Llama 2, features improved performance and is more suitable for business use.
Like other AI chatbots, Llama 2 had to be trained using online data. Exposure to this vast resource of information helps it improve what it does – providing users with useful responses to their questions.
An initial version of Llama 2 was created through “supervised fine-tuning”, a technique that uses high-quality question-and-answer data to calibrate it for use by the public. It was further refined with human feedback reinforcement learning which, as the name suggests, incorporates people’s assessments of the AI’s performance to align it with human preferences.
Guaranteed benefits Meta’s embrace of the open-source ethos with Llama 2 allows it to capitalise on what appears to be an approach that has worked for the company in the past. Meta’s engineers are known for their development of products to aid developers such as React and PyTorch. Both are open source and have become the industry standard. Through them, Meta has set a precedent of innovation through collaboration.
The release of Llama 2 holds the promise of safer generative AI. Through shared wisdom and collective exploration, users can identify erroneous information and any vulnerabilities that could be exploited by criminals. Unexpected applications have already emerged, such as a version of Llama 2 that can be installed on iPhones and was created by users, underscoring the potential for creativity within this community.
But there are limits to how far Meta will allow Llama 2 users to commercialise its AI system. If any party achieves more than 700 million active users in the preceding calendar month for a product based on Llama 2, it will have to request a licence from Meta. For Meta, this opens up the potential for profit-sharing on successful products based on Llama 2.
Meta’s strategy contrasts starkly with the more guarded approach of its primary competitor, OpenAI. Even as some question Meta’s ability to compete in this area and commercialise products as OpenAI has done with ChatGPT, Meta’s decision to invite worldwide developers into the fold suggests a broader vision. It’s a move that positions Zuckerberg’s company not merely as a player but a facilitator, harnessing global talent to contribute to the growing ecosystem of Llama 2.
This strategy could also be a shrewd hedge against potential competition from fellow tech giants such as Google. With a large population of users exploring the potential of Llama 2, any successful advance can be promptly integrated into Meta’s other products. Only time will reveal the full impact of this decision, but the immediate effects on the industry are already resonating far and wide.
Advantages and pitfalls for users The public experimentation aspect of open source technology allows for greater scrutiny, providing an opportunity for a community of users to assess Llama 2’s strengths and weaknesses, including its vulnerability to attacks. The public’s watchful eye may reveal flaws in LLMs, prompting the development of defences against them.
On the downside, concerns have emerged that this is akin to “handing a knife to criminals”, as it could also allow malicious users to exploit the technology. For example, its power could help fraudsters build a dialogue system that generates plausible automated conversations for telephone scams. This potential for misuse has led some to call for regulation of the technology.
But exactly what rules are devised, who gets the power to supervise this process, and exactly what needs more or less scrutiny, all require careful planning to make sure that regulation does not simply prop up monopolies for the big tech companies.
In the evolving saga of AI development, the debate over open sourcing serves as a reminder that technological advancements are rarely simple or one-dimensional. The implications of Meta’s decision are likely to ripple across the tech world for years to come. While Llama 2 may not yet rival the capabilities of ChatGPT, it opens the door to the development of a host of innovative products.
Google will also be under scrutiny, as speculation grows about how it may respond. In an era where open source culture thrives, it would not be surprising to see Google follow suit with its own releases.
The term “tech for good” has become a common mantra to describe technology companies using some of their resources to make a positive impact on all our lives. Ultimately, though, this objective remains a shared responsibility, not just something that a handful of companies should be engaged in.
It’s also an aim that demands collaboration and a concerted effort across academia, industry, and beyond. As LLM technologies continue to evolve, the stakes are high, and the path forward is laden with both opportunities and challenges.
Samsung launched the Galaxy Z Fold 5 and Galaxy Z Flip 5 alongside the Galaxy Tab S9 series and Galaxy Watch 6 series at its first Galaxy Unpacked event in South Korea. We discuss the company’s new devices and more on the latest episode of Orbital, the Gadgets 360 podcast. Orbital is available on Spotify, Gaana, JioSaavn, Google Podcasts, Apple Podcasts, Amazon Music and wherever you get your podcasts.
(This story has not been edited by NDTV staff and is auto-generated from a syndicated feed.)
Affiliate links may be automatically generated – see our ethics statement for details.
YouTube will roll out a slew of artificial-intelligence-powered features for creators, the latest effort from parent company Alphabet to incorporate generative AI — technology that can create and synthesize text, images, music and other media given simple prompts — into its most important products and services.
Among the new products YouTube announced Thursday is a tool called Dream Screen that uses generative AI to add video or image backgrounds to short-form videos, which the company calls Shorts. It also announced new AI-enabled production tools to help with editing both short- and long-form videos on its platform.
“We’re unveiling a suite of products and features that will enable people to push the bounds of creative expression,” Toni Reid, YouTube’s vice president for community products, said in a blog post timed to the announcement Thursday. The Google-owned video platform first announced that it was developing the tools in March.
Google has been under pressure to show results and practical applications for its generative AI products. Some critics have been wary the company, which has long been seen as a leader in artificial intelligence, was falling behind upstarts like OpenAI or rival Microsoft, and that the products Google was rolling out weren’t yet ready for public consumption. OpenAI’s ChatGPT and a new Bing chatbot from Microsoft — which has invested $13 billion (nearly Rs. 1,08,100 crore) in OpenAI since 2019 — have been wildly popular and gained mainstream favour.
Over the past few months, Google launched its own ChatGPT competitor, Bard, and released a steady flow of updates to the product. It’s also incorporated experimental generative AI features into its most important services, including its flagship search engine, in what the company calls its experimental “search generative experience.” The product generates detailed summaries based on information it’s ingested from the internet and other digital sources in response to search queries.
The announcement of the new features also comes as YouTube is locked in fierce competition with ByteDance‘s TikTok and Meta Platforms‘s Instagram Reels to gain more share of the vertical, short-form video market. YouTube said it now sees more than 70 billion daily views on Shorts, and the new generative AI tools appear to be aimed at attracting even more users and creators and gaining a competitive edge over its rivals.
The company also announced YouTube Create, a mobile app aimed at helping the platform’s creators make video production work easier. The app includes AI-enabled features like editing and trimming, automatic captioning, voiceover capabilities and access to a library of filters and royalty-free music. The app is currently in beta on Android in “select markets,” the company said, and will be free of charge.
Beyond creation, YouTube said it would also provide creators with more tools to get AI-powered insights, help with automatic dubbing of videos and assist with finding music and soundtracks for videos.
© 2023 Bloomberg LP
Affiliate links may be automatically generated – see our ethics statement for details.
WhatsApp has begun rolling out support for a new feature that will allow you to log in to your account using the biometric authentication mechanism on your smartphone. The messaging service will soon allow you to create a passkey — a kind of login credential that eliminates the need to use or remember a password — on your device and use it to securely log in to apps and services using the facial recognition or fingerprint scanner on your device.
Feature tracker WABetaInfo spotted the new passkey feature on WhatsApp beta for Android 2.23.20.4 on Tuesday, that is rolling out to beta users. However, not all users who have updated to the latest beta release will have access to the feature, which is reportedly rolling out to a “limited number of beta testers”. Gadgets 360 was unable to access the feature on two different Android smartphones that are both enrolled in the beta program.
The new Passkeys feature on WhatsApp
Photo Credit: WABetaInfo
The new passkey feature is described as a “simple way to sign in safely” to WhatsApp in a screenshot shared by the feature tracker. This suggests that it could be used to help sign in to other devices via secure authentication on your primary device.
Authenticating using passkeys isn’t a novel concept and the technology is slowly gaining traction online— Google already allows you to log in to a new device by using fingerprint-based biometric authentication for passkeys in place of a password. These passkeys are securely stored on your device and used when biometric authentication is provided.
The screenshot posted by WABetaInfo also states that WhatsApp will store the passkey in the device’s password manager — for most users, that would be the device’s default password store that is handled by Google with autofill support. The feature is also expected to make its way to iOS, where it is likely to be stored in the iOS Keychain.
It is currently unclear whether WhatsApp will also support storing passkeys in third-party apps like Bitwarden, 1Password, or Dashlane. We can expect to learn more about how the feature works when it is rolled out to more users in the beta program and the feature is expected to arrive on all smartphones on the stable channel in the future.
Affiliate links may be automatically generated – see our ethics statement for details.
Britain urged Meta not to roll out end-to-end encryption on Instagram and Facebook Messenger without safety measures to protect children from sexual abuse after the Online Safety Bill was passed by parliament.
Meta, which already encrypts messages on WhatsApp, plans to implement end-to-end encryption across Messenger and Instagram direct messages, saying the technology re-enforced safety and security.
Britain’s Home Secretary Suella Braverman said she supported strong encryption for online users but it could not come at the expense of children’s safety.
“Meta has failed to provide assurances that they will keep their platforms safe from sickening abusers,” she said. “They must develop appropriate safeguards to sit alongside their plans for end-to-end encryption.”
A Meta spokesperson said: “The overwhelming majority of Brits already rely on apps that use encryption to keep them safe from hackers, fraudsters and criminals.
“We don’t think people want us reading their private messages so have spent the last five years developing robust safety measures to prevent, detect and combat abuse while maintaining online security.”
It said it would update on Wednesday on the measures it was taking, such as restricting people over 19 from messaging teens who do not follow them and using technology to identify and take action against malicious behaviour.
“As we roll out end-to-end encryption, we expect to continue providing more reports to law enforcement than our peers due to our industry leading work on keeping people safe,” the spokesperson said.
Social media platforms will face tougher requirements to protect children from accessing harmful content when the Online Safety Bill passed by Parliament on Tuesday becomes law.
End-to-end encryption is a bone of contention between companies and the government in the new law.
Messaging platforms led by WhatsApp oppose a provision that they say could force them to break end-to-end encryption.
The government, however, has said the bill does not ban the technology, but instead, it requires companies to take action to stop child abuse and as a last resort develop technology to scan encrypted messages.
Tech companies have said scanning messages and end-to-end encryption are fundamentally incompatible.
© Thomson Reuters 2023
(This story has not been edited by NDTV staff and is auto-generated from a syndicated feed.)
Affiliate links may be automatically generated – see our ethics statement for details.
How To Save Tons of Time With Instagram Auto Reply
Instagram Emoji Guide: Meanings, Reactions, Ideas
YouTube Announces AI-Enabled Editing Products for Video Creators
WhatsApp Passkey Support Reportedly Rolling Out to Beta Testers on Android: How It Works
Career stories: Influencing engineering growth at LinkedIn
Meta Urged Not to Roll Out End-to-end Encryption on Messenger, Instagram by UK
Allowing Users to Promote Stories as Ads (via Marketing API)
![17-tips-for-writing-effective-youtube-descriptions-[free-tool]](https://www.social-ping.com/wp-content/uploads/2023/08/127190-17-tips-for-writing-effective-youtube-descriptions-free-tool-80x80.png)
17 Tips for Writing Effective YouTube Descriptions [Free Tool]
Facebook Hashtags: Free Hashtag Tool + 70 Ideas
How to Boost Engagement With LinkedIn Carousel Posts and Ads
Enhancing homepage feed relevance by harnessing the power of large corpus sparse ID embeddings
Enhancing Security and Developer Productivity: LinkedIn’s Journey with Implementing Content Security Policy
-
Uncategorized1 week ago3 Ways To Find Your Instagram Reels History
-
OTHER2 weeks agoWhatsApp Chat Interoperability Feature Spotted in Development on Latest Beta Update: Report
-
FACEBOOK1 week agoIntroducing Facebook Graph API v18.0 and Marketing API v18.0
-
Uncategorized1 week agoCommunity Manager: Job Description & Key Responsibilities
-
LINKEDIN1 week agoCareer Stories: Learning and growing through mentorship and community
-
OTHER2 weeks agoYouTube ‘Subscribe’ Button Spotted to Be Glowing When Creators Request Subscription
-
Uncategorized1 week agoThe Complete Guide to Social Media Video Specs in 2023
-
Uncategorized1 week agoSocial Media Intelligence: What It Is & Why You Need It