Connect with us

Technology

G42 launches NANDA—a new Hindi language model at UAE-India Business Forum- The Week

Published

on

G42 launches NANDA—a new Hindi language model at UAE-India Business Forum- The Week

G42, a leading AI-based firm in Abu Dhabi, UAE, has announced the upcoming launch of NANDA, a Hindi Large Language Model (LLM), designed predominantly for Hindi speakers in India.

The announcement was made in the presence of His Highness Sheikh Khaled bin Mohammed bin Zayed Al Nahyan, Crown Prince of Abu Dhabi during the UAE-India Business Forum in Mumbai on September 10, 2024.

NANDA was named after Nanda Devi, second-highest mountain in India, after Kangchenjunga. The program is a 13-billion parameter model trained on approximately 2.13 trillion tokens of language datasets, including Hindi, English and Hinglish.

The project is a collaboration between Inception (a G42 company), Mohamed bin Zayed University of Artificial Intelligence (the world’s first graduate research university dedicated to AI) and Cerebras (an American artificial intelligence company based in Sunnyvale, California, with an office in Bangalore).

Advertisement

Manu Jain, the CEO of G42 India, has said that India has solidified its position as a “global technology leader”, with many initiatives like ‘Digital India’ and ‘Startup India’ emerging under Prime Minister Narendra Modi’s leadership.

As the country stands on the brink of AI-powered growth, G42 is proud to contribute to this journey with the launch of NANDA in support of India’s AI ambitions,” says Jain.

Dr Andrew Jackson, Acting CEO of Inception, says that, “G42 has a strong track record in the development of language and domain-specific LLMs. With NANDA, we are heralding a new era of AI inclusivity, ensuring that the rich heritage and depth of Hindi language is represented in the digital and AI landscape. NANDA exemplifies G42’s unwavering commitment to excellence and fostering equitable AI.”

Earlier this year, leading software company Microsoft, invested $1.5 billion in G42, hoping to advance AI’s capabilities and broaden its reach.

Advertisement

This is not the first time that G42 has created a language model AI. Back in 2023, they launched an Arabic LLM called ‘JAIS’, the first of its kind.

“JAIS set a new standard for linguistic AI which G42 now seeks to replicate for other regions whose languages are still underrepresented,” the company said.

Building on the success of JAIS, the release of NANDA hopes to empower India’s scientific, academic, and developer communities while at the same time growing the Hindi language AI ecosystem.

Advertisement

Source link

Continue Reading
Advertisement
Click to comment

You must be logged in to post a comment Login

Leave a Reply

Technology

Razer’s Xbox quick charge stand drops to $20

Published

on

Razer's Xbox quick charge stand drops to $20

Right now Amazon has an incredible deal on the Razer Universal Quick Charge stand for Xbox controllers that saves you quite a bit of money on the accessory. This charging stand will fit one controller and it normally retails for $40 to $50 depending on the style you pick. For instance, some styles like the Electric Volt are $40 at full price while the Forza Horizon 5 model is $50 at full price.

However, Amazon is knocking the price down on many of the available options to $20. So depending on the color or style you choose, you’re saving either 50% or 60%. This is easily the lowest price we’ve tracked for this accessory at any point and if you need an easy way to keep your wireless controller charged, this is a deal you don’t want to miss.

Razer Quick Charge Stand Price History

Now it is probably obvious but it needs to be said for clarification, this deal is only for Razer’s quick charge stand and it does not come with an Xbox wireless controller. The stand does come with the necessary USB cable to plug it into a power source though. You can sit your controller on the stand at any point and charge your wireless controller and it should be charged up in under a few hours.

The stand also serves as a nice place to store the controller when you’re not playing, which also looks nice. So it keeps things organized in your setup. In terms of the available colors/styles for this deal, Amazon is offering the $20 price on the Deep Pink, Electric Volt, Forza Horizon 5, Mineral Camo, Robot White, and Velocity Green variants.

Advertisement

Buy at Amazon

Source link

Advertisement
Continue Reading

Technology

Informatica launches blueprints for developing generative AI

Published

on

Informatica launches blueprints for developing generative AI

Informatica launched Generative AI Blueprints, a set of guidelines and features designed to speed and simplify developing generative AI tools on six prominent cloud-based data management platforms.

The blueprints, available at no cost in Informatica’s Architecture and were released on Oct. 24, enable customers to use the vendor’s Intelligence Data Management Cloud (IDMC) platform in concert with AWS, Databricks, Google Cloud, Microsoft Azure, Oracle Cloud Infrastructure and Snowflake to develop generative AI capabilities.

Key features of the blueprints include vector database and language model connectors, prebuilt no-code data integration recipes, data quality and master data management capabilities and standard reference architectures to guide developers as they build new tools.

Advertisement

Given that enterprise interest in developing generative AI models and applications has exploded over the past two years yet the actual building of such tools is complicated, Informatica’s launch of Generative AI Blueprints will be useful for the vendor’s customers, according to Kevin Petrie, an analyst at BARC U.S.

“Data and AI teams struggle to integrate the many elements of a GenAI architecture, from data sources to pipelines to [retrieval-augmented generation] workflows and the applications that contain language models,” he said. “The more you can simplify this architecture, the better you can reduce risk to determine the value of pilots and limited production deployments.”

Beyond being assistive to customers, the launch of Generative AI Blueprints also represents Informatica capitalizing on an opportunity, Petrie continued.

BARC’s research shows that more than 80% of enterprises are likely or extremely likely to work with vendors when developing AI initiatives. In addition, more than half are likely or extremely likely to engage a global consulting firm.

Advertisement

“As incumbent providers in many environments, Informatica and its consulting partners recognize this opportunity and aim to capitalize on it,” Petrie said.

Based in Redwood City, Calif., Informatica is a data management vendor whose IDMC is designed to enable its users to integrate and prepare data for analysis and AI development.

Earlier this month, the vendor’s most recent platform update featured capabilities aimed at enabling customers to prepare their data for developing AI models and applications. Generative AI Blueprints builds on that by providing users with the next steps to take that data and use it to build models and applications.

New capabilities

Generative AI is being termed a transformative technology, akin to the smartphone 15 years ago, the internet in the 1990s, the personal computer in the 1980s and the telephone a century ago. Generative AI has existed for years, but OpenAI’s November 2022 launch of ChatGPT was a significant advancement in its capabilities and ease of use that led to surging interest.

Advertisement

Two of its primary benefits are true natural language processing that enables non-technical workers to use complex tools that previously required coding knowledge and other extensive training as well as making technical experts more efficient by automating time-consuming repetitive tasks.

As a result, many enterprises are developing — or at least preparing to develop — generative AI tools.

At their core is data, which is needed to train models and applications to understand an enterprise’s operations. In response, many data management and analytics vendors including Informatica are developing generative AI-powered tools such as conversational assistants to help customers use their platform. In addition, they are creating ecosystems for those customers to develop their own generative AI capabilities trained on their proprietary data.

For example, data platform vendors Databricks and Snowflake each are building up environments for their users to develop generative AI capabilities as are tech giants AWS, Google, Microsoft and Oracle.

Advertisement

However, developing generative AI — as well as traditional AI and machine learning — models and applications is difficult. It requires significant data preparation to ensure the quality of the data used to train generative AI tools, integrations with language models that provide the generative AI itself and the creation of complex pipelines that ultimately feed models and applications the data that trains them.

It’s estimated that over 80% of all AI projects never make it into production. And in July, research and advisory firm Gartner predicted that 30% of generative AI projects will be abandoned as soon as the end of 2025.

Informatica’s Generative AI Blueprints are designed to assist enterprise customers by simplifying the complex development process, providing them with not only needed tools but also expert advice.

Rik Tamm-Daniels, Informatica’s global vice president of strategic ecosystems and technology, noted that as enterprises attempt to develop generative AI tools, even knowing where to start can be a challenge. Generative AI Blueprints lay out the steps for them.

Advertisement

“[Two-thirds] of tech leaders expect their organization to invest more in AI over the next three years, so finding an efficient and effective implementation process is crucial,” he said. “Successfully implementing AI requires a robust framework to ensure that an enterprise’s data is ready.”

A mix of customer feedback and internal conversations among product teams led to the development of the blueprints, Tamm-Daniels continued.

“We realized that we could support the deployment and scaling of enterprise-grade GenAI applications by creating an easy-to-follow template that provides organization-specific end-to-end solutions,” he said.

Organizations, meanwhile, are seeking support as they attempt to build generative AI tools, according to Stephen Catanzano, an analyst at TechTarget’s Enterprise Strategy Group. Some are having early success. Many others, however, have plans to build generative AI tools but aren’t yet able.

Advertisement

As a result, Informatica’s Generative AI Blueprints address a real need.

“Organizations are looking for leadership from their technology partners on how to leverage their enterprise data to build contextual generative AI application leveraging their preferred cloud partners,” Catanzano said. “These blueprints are what organizations need to navigate this complex and ever-changing AI landscape.”

Specific features included in Informatica’s Generative AI Blueprints include the following:

  • Standard reference architectures that instruct customers how to develop a generative AI model or application and pre-defined configurations that simplify the actual construction of those models and applications.
  • Prebuilt no-code frameworks for using the IDMC’s integration and orchestration capabilities on AWS, Azure, Google and Oracle.
  • Connectors to the language models and vector databases needed as part of the development and training pipeline.
  • Data quality and master data management capabilities to help customers ensure the relevancy and quality of the data they use to train and inform generative AI tools.
  • Metadata and data governance to further improve the quality and relevancy of data underpinning generative AI models and applications.
  • Policy and security enforcement capabilities to address responsible development of AI tools by making sure the data included in prompts responds appropriately and securely.
  • A design interface that enables users to scale usage of high throughput/low latency serverless runtime when needed to integrate retrieval-augmented generation pipelines and other high volume workloads.

Combined, the features comprising Informatica’s Generative AI Blueprints are thorough, according to Catanzano.

“They look very comprehensive,” he said, noting that the different capabilities address everything from identifying a use case for AI through cultivating trusted data and vectorizing the data to make it discoverable by AI-powered automation tools, to choosing an LLM with which to use data and building the retrieval-augmented generation pipeline that feeds the model or application.

Advertisement

“It shows how important an ecosystem is for Informatica and their partners to work together,” Catanzano said. “It’s very comprehensive when you leverage all of the partners in their Blueprints.”

Petrie similarly noted that Informatica’s Generative AI Blueprints effectively address data preparation and pipeline development.

“These blueprints can help enterprises prepare, validate and deliver data for AI, especially structured tables that often contain enterprises’ most trustworthy data,” he said.

Looking ahead

Informatica’s recent platform update focused on preparing data for AI development and the launch of Generative AI Blueprints continue more than of year making AI a focal point of product development. That will continue, according to Tamm-Daniels.

Advertisement

Like many analytics and data management vendors, Informatica understood the potential of generative AI and the ways it could use the technology to develop tools to help its customers. In addition, it understood that enterprises would want to develop their own generative AI capabilities and unveiled features aimed at assisting them in that pursuit.

For example, in May 2023, Informatica unveiled plans to infuse Claire, its AI engine, with generative AI. Similarly, competitors including Alteryx and Fivetran have made AI part of their product development.

“We will continue to focus on strengthening and enhancing our GenAI roadmap and solutions as more and more enterprises look to Informatica to solve their data management problems,” Tamm-Daniels said.

One way Informatica could strengthen its Generative AI Blueprints would be to add connectors to graph databases in addition to the connectors to vector databases that are already included, according to Petrie.

Advertisement

Just as vector search and storage can aid generative AI development by enabling developers to discover relevant data, graph technology can similarly help the data discovery process.

“I see much less focus on graph elements in these architectures,” Petrie said. “Knowledge graphs matter because they … help GenAI language models understand how concepts and business entities relate to one another. I’ll be interested to see how Informatica addresses this requirement in coming releases.”

Catanzano, meanwhile, suggested that Informatica build on Generative AI Blueprints by providing even more guidance to its customers. Providing tools for development is essential. But so is guiding enterprises through the use of those tools to get their desired results.

“Continuing thought leadership in this area is important,” Catanzano said. “Everyone is still trying to figure it out and they need their tech partners to help guide them.”

Advertisement

Eric Avidon is a senior news writer for TechTarget Editorial and a journalist with more than 25 years of experience. He covers analytics and data management.

Source link

Continue Reading

Technology

The UK’s antitrust regulator will formally investigate Alphabet’s $2.3 billion Anthropic investment

Published

on

Menu

The UK’s competition regulator is probing Alphabet’s investment in AI startup Anthropic. After opening public comments this summer, the Competition and Market Authority (CMA) said on Thursday it has “sufficient information” to begin an initial investigation into whether Alphabet’s reported $2.3 billion investment in the Claude AI chatbot maker harms competition in UK markets.

The CMA breaks its merger probes into two stages: a preliminary scan to determine whether there’s enough evidence to dig deeper and an optional second phase where the government gathers as much evidence as possible. After the second stage, it ultimately decides on a regulatory outcome.

The probe will formally kick off on Friday. By December 19, the CMA will choose whether to move to a phase 2 investigation.

Google told Engadget that Anthropic isn’t locked into its cloud services. “Google is committed to building the most open and innovative AI ecosystem in the world,” a company spokesperson wrote in an email. “Anthropic is free to use multiple cloud providers and does, and we don’t demand exclusive tech rights.” Engadget also reached out to the CMA for comment, and we’ll update this story if we hear back.

Advertisement

TechCrunch notes that Alphabet reportedly invested $300 million in Anthropic in early 2023. Later that year, it was said to back the AI startup with an additional $2 billion. Situations like this can be classified as a “quasi-merger,” where deep-pocketed tech companies essentially take control of emerging startups through strategic investments and hiring founders and technical workers.

Amazon has invested even more in Anthropic: a whopping $4 billion. After an initial public comment period, the CMA declined to investigate that investment last month. The CMA said Amazon avoided Alphabet’s fate at least in part because of its current rules: Anthropic’s UK turnover didn’t exceed £70 million, and the two parties didn’t combine to account for 25 percent or more of the region’s supply (in this case, AI LLMs and chatbots).

Although the CMA hasn’t specified, something in Alphabet’s $2.3 billion Anthropic investment constituted a deeper dive. Of course, Google’s Gemini competes with Claude, and both companies make large language models they provide to small businesses and enterprise customers.

Update, October 25, 2024, 11:10AM ET: This story has been updated to add a quote from a Google representative.

Advertisement

Source link

Continue Reading

Technology

OpenAI CEO calls GPT-5 Orion report ‘fake news out of control’

Published

on

OpenAI CEO calls GPT-5 Orion report 'fake news out of control'

Join our daily and weekly newsletters for the latest updates and exclusive content on industry-leading AI coverage. Learn More


The Verge last night published an exclusive and seemingly well researched and sourced report (it’s great in my opinion, read it here) from journalists Kylie Robison and Tom Warren stating that OpenAI plans to launch another new frontier AI model, codenamed Orion — which may or may not be GPT-5 — by December.

Yet two hours after the article went live, Sam Altman, OpenAI’s co-founder and CEO, took to X to respond by replying directly to Robison’s share of the article, writing “fake news out of control.”

Altman hasn’t elaborated much since then from what I’ve seen, and the response is notably not exactly a direct denial of the claims — he didn’t write “No” or “this is false,” much less describe which part of the detailed article is wrong: is OpenAI not working on a new frontier model called Orion? That would contradict prior reporting from outlets including The Information that it does have such an effort internally — which to my knowledge, OpenAI never directly denied. Is it not planning to release later this year?

Advertisement

But it is clearly an attempt to push back on the reporting as it stands.

It’s an interesting quasi-denial given how precise The Verge report is, noting specific details about Orion’s supposed release plans and the fact that it appears to be geared toward enterprise customers and possibly would be served up through an application programming interface (API) only at first:

“Unlike the release of OpenAI’s last two models, GPT-4o and o1, Orion won’t initially be released widely through ChatGPT. Instead, OpenAI is planning to grant access first to companies it works closely with in order for them to build their own products and features, according to a source familiar with the plan.

Another source tells The Verge that engineers inside Microsoft — OpenAI’s main partner for deploying AI models — are preparing to host Orion on Azure as early as November. While Orion is seen inside OpenAI as the successor to GPT-4, it’s unclear if the company will call it GPT-5 externally.

Advertisement

OpenAI’s last release of a new frontier modelo1 preview and o1-mini — occurred in early September, a little more than a month ago. Yet the wider reception of these large language models (LLMs) has been largely muted, in part because they are expensive for both the company and developers to operate, and also because they are of a new “reasoning” architecture and are more limited in many ways than OpenAI’s GPT family of models, unable at this time to accept file uploads, or to generate and analyze imagery.

A new frontier model would help OpenAI capture the limelight again from rivals including Anthropic, who just this week unveiled a promising new agentic mode called “Computer Use” and new version of its Claude family of LLMs. OpenAI is not in ppor

Whether OpenAI does end up releasing a new frontier model later this year or not, we’ll be following closely. For now, it seems, fans of the company and its models shouldn’t get their hopes up too soon.


Source link
Continue Reading

Technology

YC startup Pharos lands a $5M seed led by Felicis to bring AI to hospital quality reporting

Published

on

Pharos

Medical and administrative staff are increasingly overwhelmed with piles of paperwork they have to fill out every day.

Dozens, if not hundreds, of startups, are seeing opportunities to make those bureaucratic processes less burdensome with the help of generative AI. These companies are building AI medical scribes, platforms for pre-authorizing health insurance payments, and products for automatically extracting medical coding from patients’ electronic medical records (EMRs.)

But Pharos, a company that was a part of Y Combinator’s summer 2024 cohort, is applying AI to tackle another somewhat under-the-radar administrative function for hospitals: quality reporting to external clinical registries.  

Organizations like Centers for Medicare & Medicaid Services (CMS) and the American College of Surgeons aim to measure each healthcare centers’ record on delivering safe and effective care for patients. Although reporting to these registries is not always mandatory, it’s often in the best interest of hospitals. These external organizations play a crucial role in identifying quality issues (such as an increase in post-surgery infections), which can be addressed to improve patient care.

Advertisement

However, reporting to the registries is extremely time-consuming. Nurses and other staff must manually sift through each patient’s electronic health record to extract the precise data required for each registry. “A single case can take up to eight hours” to report, said Ryan Isono, a partner at Felicis, “It’s a big problem, but one that you only know about if you’re deep in the industry.”

Indeed, Pharos was co-founded by Felix Brann and Matthew Jones, who had some exposure to the challenges of reporting data to medical registries from their prior work at Vital, a startup that develops software for emergency rooms. They recognized that AI can take unstructured data from EMRs and automatically populate forms required by registries. As they went through YC earlier this year, they added another co-founder – Alex Clarke, a medical doctor who also holds a PhD in artificial intelligence from Imperial College London.

On Friday, Pharos announced that Felicis, with participation from General Catalyst, Moxxie and Y Combinator, led its $5 million seed round.

Pharos caught Felicis’ eye not only because the company could save hospitals money and free up nurses’ time for taking care of patients, but also because the area still doesn’t have other startups going after it, Isono said.

Advertisement

Brann (pictured center above) predicts that other quality reporting companies will emerge soon. “We have five years of experience selling and deploying into hospitals, and we have top-tier AI talent,” he said.” That Venn diagram doesn’t normally overlap. That’s why we think we’re going to win.”

For now, the entire Pharos team consists only of the three co-founders, but they will be using the capital to hire a team that will help the company sell the product and maintain relationships with hospitals. 

Source link

Continue Reading

Technology

UnitedHealth data breach leaked info on over 100 million people

Published

on

The Internet Archive hackers still have access to its internal emailing tools

On February 12, criminals used compromised credentials to remotely access a Change Healthcare Citrix portal, an application used to enable remote access to desktops. The portal did not have multi-factor authentication. Once the threat actor gained access, they moved laterally within the systems in more sophisticated ways and exfiltrated data. Ransomware was deployed nine days later.

Source link

Continue Reading

Trending

Copyright © 2024 WordupNews.com