Connect with us

Tech

Another ChatGPT rival lands as Meta releases Llama 2

Published

on

Facebook’s parent company, Meta Platforms, has developed an artificial intelligence (AI) system named Llama 2, which rivals ChatGPT and Google’s Bard, but with a unique twist – it will be released for free. 

The move by Meta allows startups and businesses to compete with OpenAI’s ChatGPT and Google’s Bard at a lower cost. Microsoft will distribute Llama 2 through its Azure cloud service, with Meta referring to Microsoft as its “preferred partner” for the release. 

Llama 2 is part of Meta’s large language model (LLM) series, which underpins generative AI products like ChatGPT. 

Meta CEO Mark Zuckerberg announced that the technology would be available for both research and commercial use, emphasising the company’s commitment to openness and innovation.

Meta has aimed to differentiate itself from other tech giants by being more transparent in sharing the data and code it uses to develop AI systems. Zuckerberg believes that open-source models encourage innovation and improve safety and security. In line with this philosophy, Meta has open-sourced Llama 2 and highlighted its history of open-sourcing AI work, including the widely used machine-learning framework PyTorch.

However, the research paper introducing Llama 2 is less transparent than Meta’s previous work, omitting details about the specific data used for training. The model was trained on a combination of publicly available data, excluding Meta’s own products or services. Notably, data from websites containing significant personal information was removed.

Meta will make its AI models available for download directly or through Microsoft’s Azure cloud platform. Financial terms of the partnership were not disclosed, but Meta confirmed that the models would also be accessible via Amazon Web Services, Hugging Face, and other platforms. While Microsoft is labeled as a “preferred” partner, it is worth noting that Microsoft is also a major funder and partner of OpenAI, the creator of ChatGPT.

In addition to the Llama 2 announcement, Microsoft revealed at its Inspire event that it would charge businesses a monthly fee of $30 per user for its generative AI tool, Microsoft 365 Copilot.

Latest News

Bitcoin makes its first breach of $94,000.

Published

on

By

As news broke that Donald Trump’s social media company was in negotiations to acquire cryptocurrency trading platform Bakkt (BKKT.N), Bitcoin surged to a record high above $94,000, adding to optimism that the next Trump administration would be cryptocurrency-friendly.

This year, Bitcoin, the largest and most well-known cryptocurrency in the world, has more than doubled in value. Just before the close of the previous session, it reached a record high of $94,078 before closing at $92,104 in Asian hours on Wednesday.

According to the Financial Times, which cited two persons with knowledge, NYSE-owner Intercontinental Exchange (ICE.N) is supporting Trump Media and Technology Group (DJT.O), opens new tab, which runs Truth Social, in its pursuit of an all-stock purchase of Bakkt.

In addition to traders taking advantage of the first day of options trading on the Nasdaq over BlackRock’s Bitcoin ETF (IBIT.O), Tony Sycamore, market analyst at IG, stated that the Trump deal chatter had helped push bitcoin to a record high.

As traders hoped that President-elect Trump’s pledged support for digital assets would result in a less stringent regulatory framework and give bitcoin a boost after a sluggish few months, cryptocurrencies have surged since the U.S. election on November 5.

According to data and analytics aggregator CoinGecko, the global cryptocurrency market is now worth over $3 trillion, a new high, due to the growing excitement.

Australian online broker Pepperstone’s head of research, Chris Weston, stated that there is genuine underlying purchasing pressure on bitcoin and that “another kick higher should bring in a fresh chase from those who like to buy what’s strong.”

Continue Reading

Latest News

The PTA discloses how many VPNs are registered in Pakistan.

Published

on

By

Speaking at the Senate Standing Committee on IT meeting in Islamabad, which was presided over by Palwasha Khan, was the PTA chief.

The continuous regulation of Virtual Private Networks (VPNs) in Pakistan was the main topic of discussion. The committee called for the next meeting and voiced concerns about the Minister of State for IT’s absence for the third consecutive session.

“We have spent the last two years developing VPN legislation. Uninterrupted internet connection is guaranteed by registered VPNs without sacrificing privacy, according to the PTA chairman.

During the session, Senator Muhammad Humayun criticized the measures, saying, “Shutting down platforms like X (formerly Twitter) is jeopardizing the entire IT industry and impacting the livelihoods of over 2.5 million individuals in Pakistan.”

He opposed laws that would hurt independent contractors and small companies.

Although the PTA promised that registered VPNs would not experience internet outages, committee members wanted to know the exact legal foundation for these regulations. Senator Kamran Murtaza asked for documents that detailed the requirements for registration.

“If VPN services are abruptly halted, there will be a major backlash,” Senator Afnan Ullah said. While pointing out that Pakistan’s IT exports had increased by 20% in the previous year, he emphasised that the sector was still lagging behind India.

Concerns over the viability of the existing VPN whitelist method were voiced by the committee. While IT officials admitted that more than 1 million freelancers in Pakistan rely significantly on VPNs for their business, Senator Humayun pointed out that the laws should be more beneficial than harmful.

Continue Reading

Latest News

Pakistan declares AI chatbots to be dangers to security.

Published

on

By

The National Computer Emergency Response Team (CERT) has released a security advisory concerning the increasing utilization of artificial intelligence (AI) chatbots, emphasizing potential hazards related to the exposing of private data.

The recommendation recognizes that AI chatbots, like ChatGPT, have gained significant popularity for personal and professional duties owing to their capacity to improve productivity and engagement. Nonetheless, the CERT cautions that these AI systems frequently retain sensitive information, so posing a danger of data breaches.

Engagements with AI chatbots may encompass sensitive information, such as corporate strategy, personal dialogues, or confidential correspondence, which could be compromised if inadequately safeguarded. The warning emphasizes the necessity for a comprehensive cybersecurity framework to alleviate concerns associated with AI chatbot utilization.

Users are advised against inputting critical information into AI chatbots and are encouraged to deactivate any chat-saving functionalities to mitigate the danger of unwanted data access. The CERT additionally advises performing routine system security checks and employing monitoring tools to identify any anomalous behavior from AI chatbots.

Organizations are urged to adopt rigorous security protocols to safeguard against possible data breaches resulting from AI-driven interactions.

Continue Reading

Trending