Artificial intelligence (AI), once thought to be the exclusive domain of big cities and tech hubs, is making its way to the furthest corners of rural Ireland, empowering locals with cutting-edge tools to boost their businesses and preserve their traditional crafts.

It is all part of Microsoft’s ambitious ‘Skill Up Ireland’ initiative, which aims to provide every person in Ireland with the opportunity to learn AI skills. The program has partnered with the Irish government and various organisations to deliver AI training and resources to communities across the country, leaving no one behind in the era of rapid technological advancement.

One recent beneficiary of this program is Andrew Murray, the general manager of the 22-room Doonmore Hotel on Inishbofin. A native of the island, Murray comes from a family that has lived on Inishbofin for generations, with his parents founding the hotel in 1969. Despite the remote location, Murray is eager to embrace AI as a tool to streamline his operations and save time.

“What I’m interested in the most is the power of AI to save time for people like me,” Murray said. “Because time is the most precious thing we have, and it’s finite. There are only 24 hours in a day.”

Through an AI introduction class, Murray discovered the possibilities of tools such as Microsoft Copilot, an AI-powered assistant for everything from scheduling to data analysis to creating content. He intends to use these tools to oversee things like scheduling staff and inventory management as well as invoicing and pricing – tasks that he has normally spent hours, if not days, doing completely manually.

But Murray is not alone in his enthusiasm for AI on Inishbofin. Catherine O’Connor, a weaver who draws inspiration from the island’s natural colors and textures, has also embraced the technology. Initially wary of the AI training, O’Connor quickly became “absorbed by it” once she realised its potential to help her market her handmade scarves, table runners, and wall hangings.

“Every piece has a story behind it,” O’Connor explained. “You can get a scarf at the five-and-dime store, but a handmade scarf takes hours and hours to make. It’s a totally different level. So you have to find the right words to use.”

Now, with the help of Copilot, O’Connor can write engaging descriptions of her creations for marketing her craft on a proper e-commerce platform and help people understand her work more accurately and visualise the creation.

Another Copilot user, Inishbofin-based florist Patricia Concannon, plans to also use Copilot to make her website and Instagram captions more engaging which should prove useful in helping her reach new customers and attracting a wider audience for her floral displays.

The AI training on Inishbofin is just one element of Microsoft’s wider ‘Skill Up Ireland’ programme aimed at upskill and reskill over in Ireland, which includes Dream Space, an immersive learning experience to introduce STEM and AI skills to every one of the country’s one million students and their teachers.

Kevin Marshall, head of Learning & Skills for Microsoft Ireland, said the rapid growth in the prevalence of AI in the last few years has necessitated upskilling and reskilling programmes. He continued: “At the same time, with the explosion of generative AI in the last 18 months, there’s a real need to educate people on what this is, to show them that it’s not black magic.

The challenge, however, lies in the ever-evolving nature of AI technology. “The teaching is non-invasive, it’s collaborative,” Marshall explained. “The programs teach the basic foundations and core principles of AI. Here’s what it can do. Here are the risks and the ethical issues. Here are the opportunities. And here’s where you go play with it.”

Programmes like ‘Skill Up Ireland’ are an opportunity for rural communities like Inishbofin not to be left behind through the digital divide as AI significantly impacts industries and the way that we live and work. Audrey Murray, a felt artist and teaching assistant on the island, summed it up: “AI has to be another step, I suppose, bringing us closer to the world and bringing the world here.”

And with Microsoft’s promise of creating AI skills for all in Ireland, the remote extremities of the Emerald Isle are on the brink of being catapulted into the future, when the very latest technologies are melded with ancient skills and lifeways. Meanwhile, for the inhabitants of Inishbofin, the opportunities are yet to reveal themselves.

Want to learn more about AI and big data from industry leaders? Check out AI & Big Data Expo taking place in Amsterdam, California, and London. The comprehensive event is co-located with other leading events including Intelligent Automation Conference, BlockX, Digital Transformation Week, and Cyber Security & Cloud Expo.

Explore other upcoming enterprise technology events and webinars powered by TechForge here.

The post AI comes to Ireland’s remote Islands through Microsoft’s ‘Skill Up’ program appeared first on AI News.

This new category not only boosts Microsoft’s ventures but also positions it to compete directly with AI-led advancements from tech giants such as Alphabet and Apple. During an event at Microsoft’s campus in Redmond, Washington, CEO Satya Nadella unveiled the Copilot+ PCs. These computers, developed in collaboration with manufacturers like Acer and Asustek Computer, align with Microsoft’s stock prices reaching near all-time highs amid Wall Street’s expectations that AI will significantly boost earnings for major tech firms.

These PCs are engineered to manage a greater share of AI tasks locally, reducing reliance on cloud data centres. Priced from $1,000, the devices are slated to begin shipping on June 18. The lineup includes laptops from established manufacturers such as Dell Technologies, HP, and Samsung Electronics, complemented by two models from Microsoft itself.

Technical innovations and product range

The Surface Laptop, Microsoft’s inaugural Copilot+ device, features a Qualcomm Snapdragon X series system-on-chip. This chip includes a built-in AI accelerator capable of delivering up to 45 tera operations per second (TOPS) of computing power. Available in 13.5-inch and 15-inch models, these laptops offer up to 64 gigabytes of RAM and 1 terabyte of flash storage.

Joining the Surface Laptop is the new Copilot+ version of the Surface Pro, a versatile two-in-one PC. This 10.6-inch tablet converts into a laptop with a detachable keyboard, known as the Flex Keyboard, which also functions as a wireless connector and stylus charger for the Surface Pro.

Both the Surface Laptop and Surface Pro are equipped with the Snapdragon X processor, available with up to 32 gigabytes of RAM and one terabyte of storage. The high-end models feature an OLED display, enhancing visual quality significantly.

A distinctive feature of the Copilot+ series is “Recall,” which allows the device to log all user activities—from web browsing to voice chats—and store this data locally. This functionality creates a searchable history that users can access to revisit past activities, even months later.

Yusuf Mehdi, head of consumer marketing at Microsoft, announced at the press event that the company anticipates selling 50 million AI PCs in the coming year. He emphasised that the enhanced capabilities of AI assistants, now operating directly on these PCs, will make a compelling case for consumers considering upgrades.

Despite a 15% decline in global PC shipments last year, according to Gartner, Microsoft is optimistic about the potential market share for its Copilot+ PCs.

Analyst Ben Bajarin from Creative Strategies remarked: “People just need to be convinced that the device experience alone justifies this entirely new category of Copilot+ machines.”

Echoing Intel’s 2011 strategy with the “Ultrabook,” which aimed to compete against Apple’s MacBook Air, Microsoft’s introduction of the “Copilot+” category is poised to redefine the slim-form Windows laptop market.

Microsoft also revealed that GPT-4o, the latest advancement from OpenAI’s ChatGPT, will soon be integrated into Copilot. Additionally, the new series of Surface Pro tablets and Surface Laptops will incorporate Qualcomm chips based on Arm Holdings’ designs, utilising the innovative Prism technology to adapt software originally designed for Intel and AMD processors for use on Arm-based systems.

At a recent demonstration, Microsoft showcased how its new devices, running Adobe photo editing software, outperformed an Apple device, just weeks after Apple introduced a new AI-focused chip expected to enhance future laptops.

Challenges and opportunities in the evolving PC market

Facing strong competition in the traditional PC market, Microsoft continues to lead in integrating the latest AI trends into consumer products. This strategic initiative, fueled by its partnership with OpenAI, aims to secure a leadership position in the fiercely competitive AI industry, challenging rivals like Alphabet.

Just last week, both OpenAI and Alphabet’s Google introduced new AI technologies that support real-time, voice-responsive interactions, establishing new benchmarks for AI voice assistants. Meanwhile, Windows PC manufacturers are contending with increased pressure from Apple, whose custom Arm-based chips have set new standards in battery life and performance.

Want to learn more about AI and big data from industry leaders? Check out AI & Big Data Expo taking place in Amsterdam, California, and London. The comprehensive event is co-located with other leading events including Intelligent Automation Conference, BlockX, Digital Transformation Week, and Cyber Security & Cloud Expo.

Explore other upcoming enterprise technology events and webinars powered by TechForge here.

The post Microsoft unveils new Copilot+ PCs featuring advanced AI capabilities appeared first on AI News.

According to two sources familiar with the matter who spoke to Reuters, OpenAI is scheduled to unveil its AI-powered search tool on Monday. The introduction of this product coincides closely with the timing of Google’s annual I/O conference, which starts on Tuesday and is anticipated to feature several AI-related announcements.

While Google remains a titan in search, it too has been adapting to incorporate AI more deeply into its functionalities. Google began trialling AI-powered search summaries in the US over a year ago and expanded testing to selected, signed-in UK users last month.

The potential entry of OpenAI into the search engine market adds a new layer of competition not only for Google, but also for other entities like Perplexity. This week, Perplexity announced a partnership with SoundHound to enhance voice assistants for IoT devices with its AI-powered search capabilities.

OpenAI, heavily supported by Microsoft, seems to be banking on the intersection of AI and search engine technology to carve out its niche. Microsoft previously integrated OpenAI’s AI capabilities into its Bing search engine and Edge browser in February 2023, making these advanced features available to paid subscribers, along with integration into its Microsoft Office suite.

OpenAI’s forthcoming search product is expected to expand on its existing flagship tool, ChatGPT. Reports from Bloomberg reveal that the new feature will allow users to pose questions to ChatGPT and receive responses that pull information from the web with appropriate citations. This could include references to Wikipedia or various blogs and might even provide illustrative images or diagrams relevant to the queries.

If OpenAI successfully launches this new search product, it could signify a significant shift in how internet searches are conducted—moving away from traditional search models towards more integrated, AI-driven responses that significantly enhance user experience.

(Photo by GR Stocks)

See also: Coalition of news publishers sue Microsoft and OpenAI

Want to learn more about AI and big data from industry leaders? Check out AI & Big Data Expo taking place in Amsterdam, California, and London. The comprehensive event is co-located with other leading events including Intelligent Automation Conference, BlockX, Digital Transformation Week, and Cyber Security & Cloud Expo.

Explore other upcoming enterprise technology events and webinars powered by TechForge here.

The post OpenAI set to unveil AI-driven challenger to Google Search appeared first on AI News.

First reported by The Verge, the group of eight publications owned by Alden Global Capital (AGC) – including the Chicago Tribune, New York Daily News, and Orlando Sentinel – allege the companies have purloined “millions” of their articles without permission and without payment “to fuel the commercialisation of their generative artificial intelligence products, including ChatGPT and Copilot.”

The lawsuit is the latest legal action taken against Microsoft and OpenAI over their alleged misuse of copyrighted content to build large language models (LLMs) that power AI technologies like ChatGPT. In the complaint, the AGC publications claim the companies’ chatbots can reproduce their articles verbatim shortly after publication, without providing prominent links back to the original sources.

“This lawsuit is not a battle between new technology and old technology. It is not a battle between a thriving industry and an industry in transition. It is most surely not a battle to resolve the phalanx of social, political, moral, and economic issues that GenAI raises,” the complaint reads.

“This lawsuit is about how Microsoft and OpenAI are not entitled to use copyrighted newspaper content to build their new trillion-dollar enterprises without paying for that content.”

The plaintiffs also accuse the AI models of “hallucinations,” attributing inaccurate reporting to their publications. They reference OpenAI’s previous admission that it would be “impossible” to train today’s leading AI models without using copyrighted materials.

The allegations echo those made by The New York Times in a separate lawsuit filed last year. The Times claimed Microsoft and OpenAI used almost a century’s worth of copyrighted content to allow their AI to mimic its expressive style without a licensing agreement.

In seeking to dismiss key parts of the Times’ lawsuit, Microsoft accused the paper of “doomsday futurology” by suggesting generative AI could threaten independent journalism.

The AGC publications argue that OpenAI, now valued at $90 billion after becoming a for-profit company, and Microsoft – which has seen hundreds of billions of dollars added to its market value from ChatGPT and Copilot – are profiting from the unauthorised use of copyrighted works.

The news publishers are seeking unspecified damages and an order for Microsoft and OpenAI to destroy any GPT and LLM models utilising their copyrighted content.

Earlier this week, OpenAI signed a licensing partnership with The Financial Times to lawfully integrate the newspaper’s journalism. However, the latest lawsuit from AGC highlights the growing tensions between tech companies developing generative AI and content creators concerned about the unchecked use of their works to train profitable AI systems.

(Photo by Wesley Tingey)

See also: OpenAI faces complaint over fictional outputs

Want to learn more about AI and big data from industry leaders? Check out AI & Big Data Expo taking place in Amsterdam, California, and London. The comprehensive event is co-located with other leading events including BlockX, Digital Transformation Week, and Cyber Security & Cloud Expo.

Explore other upcoming enterprise technology events and webinars powered by TechForge here.

The post Coalition of news publishers sue Microsoft and OpenAI appeared first on AI News.

“What we’re going to start to see is not a shift from large to small, but a shift from a singular category of models to a portfolio of models where customers get the ability to make a decision on what is the best model for their scenario,” said Sonali Yadav, Principal Product Manager for Generative AI at Microsoft.

The first Phi-3 model, Phi-3-mini at 3.8 billion parameters, is now publicly available in Azure AI Model Catalog, Hugging Face, Ollama, and as an NVIDIA NIM microservice. Despite its compact size, Phi-3-mini outperforms models twice its size. Additional Phi-3 models like Phi-3-small (7B parameters) and Phi-3-medium (14B parameters) will follow soon.

“Some customers may only need small models, some will need big models and many are going to want to combine both in a variety of ways,” said Luis Vargas, Microsoft VP of AI.

The key advantage of SLMs is their smaller size enabling on-device deployment for low-latency AI experiences without network connectivity. Potential use cases include smart sensors, cameras, farming equipment, and more. Privacy is another benefit by keeping data on the device.

Large language models (LLMs) excel at complex reasoning over vast datasets—strengths suited to applications like drug discovery by understanding interactions across scientific literature. However, SLMs offer a compelling alternative for simpler query answering, summarisation, content generation, and the like.

“Rather than chasing ever-larger models, Microsoft is developing tools with more carefully curated data and specialised training,” commented Victor Botev, CTO and Co-Founder of Iris.ai.

“This allows for improved performance and reasoning abilities without the massive computational costs of models with trillions of parameters. Fulfilling this promise would mean tearing down a huge adoption barrier for businesses looking for AI solutions.”

Breakthrough training technique

What enabled Microsoft’s SLM quality leap was an innovative data filtering and generation approach inspired by bedtime story books.

“Instead of training on just raw web data, why don’t you look for data which is of extremely high quality?” asked Sebastien Bubeck, Microsoft VP leading SLM research.  

Ronen Eldan’s nightly reading routine with his daughter sparked the idea to generate a ‘TinyStories’ dataset of millions of simple narratives created by prompting a large model with combinations of words a 4-year-old would know. Remarkably, a 10M parameter model trained on TinyStories could generate fluent stories with perfect grammar.

Building on that early success, the team procured high-quality web data vetted for educational value to create the ‘CodeTextbook’ dataset. This was synthesised through rounds of prompting, generation, and filtering by both humans and large AI models.

“A lot of care goes into producing these synthetic data,” Bubeck said. “We don’t take everything that we produce.”

The high-quality training data proved transformative. “Because it’s reading from textbook-like material…you make the task of the language model to read and understand this material much easier,” Bubeck explained.

Mitigating AI safety risks

Despite the thoughtful data curation, Microsoft emphasises applying additional safety practices to the Phi-3 release mirroring its standard processes for all generative AI models.

“As with all generative AI model releases, Microsoft’s product and responsible AI teams used a multi-layered approach to manage and mitigate risks in developing Phi-3 models,” a blog post stated.  

This included further training examples to reinforce expected behaviours, assessments to identify vulnerabilities through red-teaming, and offering Azure AI tools for customers to build trustworthy applications atop Phi-3.

(Photo by Tadas Sar)

See also: Microsoft to forge AI partnerships with South Korean tech leaders

Want to learn more about AI and big data from industry leaders? Check out AI & Big Data Expo taking place in Amsterdam, California, and London. The comprehensive event is co-located with other leading events including BlockX, Digital Transformation Week, and Cyber Security & Cloud Expo.

Explore other upcoming enterprise technology events and webinars powered by TechForge here.

The post Microsoft unveils Phi-3 family of compact language models appeared first on AI News.

The high-level meeting, dubbed the MS CEO Summit 2024, will be held on 14 May 2024 and feature Microsoft’s founder Bill Gates and Chairman and CEO Satya Nadella. They will engage in closed-door discussions with Kyung Kye-hyun of Samsung, Kwak Noh-jung of SK Hynix, Cho Joo-wan of LG Electronics, and Ryu Young-sang of SK Telecom.

Sources for The Korea Economic Daily suggest that Microsoft plans to explore joint ventures in AI technology across various sectors. Discussions with Samsung and SK Hynix will likely centre on the joint development and supply of AI chips.

Samsung and SK Hynix are recognised as being among the world’s leading memory chipmakers and can enhance Microsoft’s server capabilities with next-generation technologies such as High-Bandwidth Memory (HBM) AI chips and solid-state drives (SSDs).

Collaboration topics with LG Electronics will include integrating AI technologies into home appliances, a move that will boost Microsoft’s competitive edge against rivals like Google and Meta. With SK Telecom, Microsoft is expected to delve further into cloud and 5G services.

These meetings are timely, as the global tech landscape sees an increased focus on AI development. By potentially integrating Microsoft’s AI services into products like Samsung’s smartphones and LG’s home appliances, Microsoft could significantly elevate its market standing.

Kyung of Samsung’s Device Solutions indicated last month that their new AI accelerators, Mach-1 and Mach-2, will soon move into mass production. These accelerators are designed to optimise the synergy between GPUs and HBM chips, promising a revolution in processing speeds. Earlier this month, the company unveiled the industry’s first LPDDR5X DRAM which aims to boost on-device AI.

SK Telecom, under CEO Ryu, spearheads the Global Telco AI Alliance (GTAA). This consortium, including major global players like Deutsche Telekom and SingTel, aims to develop AI infrastructure and generative AI services across a customer base exceeding 1.3 billion globally.

Last year, SK Telecom invested $100 million in AI startup Anthropic to develop a large language model (LLM) specifically for telcos. The collaborative endeavour extends to the Telco AI Platform, an ongoing project initiated by the GTAA.

The MS CEO Summit 2024 presents an opportunity for enhanced AI cooperation and technological advancement, securing Microsoft’s position as a pivotal player in the industry.

(Photo by Natalie Pedigo)

See also: Meta raises the bar with open source Llama 3 LLM

Want to learn more about AI and big data from industry leaders? Check out AI & Big Data Expo taking place in Amsterdam, California, and London. The comprehensive event is co-located with other leading events including BlockX, Digital Transformation Week, and Cyber Security & Cloud Expo.

Explore other upcoming enterprise technology events and webinars powered by TechForge here.

The post Microsoft to forge AI partnerships with South Korean tech leaders appeared first on AI News.

The hub will be led by AI scientist and engineer Jordan Hoffmann, who previously distinguished himself as a pioneer at AI companies Inflection and DeepMind, which are also based in London.

“There is an enormous pool of AI talent and expertise in the UK, and Microsoft AI plans to make a significant, long-term investment in the region as we begin hiring the best AI scientists and engineers into this new AI hub,” said Mustafa Suleyman, EVP and CEO of Microsoft AI.

Suleyman, who co-founded AI startup Inflection before it was acquired by Microsoft, is a British citizen born and raised in London. “I’m proud to have co-founded and built a cutting-edge AI business here. I’m deeply aware of the extraordinary talent pool and AI ecosystem in the UK,” he commented.

Microsoft has an existing AI research presence in the UK through its Microsoft Research lab in Cambridge. However, the new dedicated Microsoft AI London hub signals the company’s increased commitment to advancing the field in Britain.

The investment builds upon Microsoft’s recently announced £2.5 billion pledge to upskill the UK workforce and build AI infrastructure – including bringing 20,000 advanced GPUs to the country by 2026.

Microsoft AI London will collaborate closely with other teams across Microsoft and partners like OpenAI. The company expects to post job openings in the coming weeks and is seeking “exceptional individuals” passionate about tackling AI’s biggest challenges.

“I know – through my close work with thought leaders in the UK government, business community, and academia – that the country is committed to advancing AI responsibly and with a safety-first commitment to drive investment, innovation, and economic growth,” said Suleyman.

“Our decision to open this hub in the UK reflects this ambition.”

(Photo by Charles Postiaux)

See also: Microsoft: China plans to disrupt elections with AI-generated disinformation

Want to learn more about AI and big data from industry leaders? Check out AI & Big Data Expo taking place in Amsterdam, California, and London. The comprehensive event is co-located with other leading events including BlockX, Digital Transformation Week, and Cyber Security & Cloud Expo.

Explore other upcoming enterprise technology events and webinars powered by TechForge here.

The post Microsoft AI opens London hub to access ‘enormous pool’ of talent appeared first on AI News.

Mistral AI stated that Mistral Large outscored most major LLMs except for OpenAI’s recently launched GPT-4 in tests of language understanding. It also performed strongly in maths and coding assessments.

Co-founder and Chief Scientist Guillaume Lample said Mistral Large represents a major advance over earlier Mistral models. The company also launched a chatbot interface named Le Chat to allow users to interact with the system, similar to ChatGPT.  

The proprietary model boasts fluency in English, French, Spanish, German, and Italian, with a vocabulary exceeding 20,000 words. While Mistral’s first model was open-source, Mistral Large’s code remains closed like systems from OpenAI and other firms.  

Mistral AI received nearly $500 million in funding late last year from backers such as Nvidia and Andreessen Horowitz. It also recently partnered with Microsoft to provide access to Mistral Large through Azure cloud services.  

Microsoft’s investment of €15 million into Mistral AI is set to face scrutiny from European Union regulators who are already analysing the tech giant’s ties to OpenAI, maker of market-leading models like GPT-3 and GPT-4. The European Commission said Tuesday it will review Microsoft’s deal with Mistral, which could lead to a formal probe jeopardising the partnership.

Microsoft has focused most of its AI efforts on OpenAI, having invested around $13 billion into the California company. Those links are now also under review in both the EU and UK for potential anti-competitive concerns. 

Pricing for the Mistral Large model starts at $8 per million tokens of input and $24 per million output tokens. The system will leverage Azure’s computing infrastructure for training and deployment needs as Mistral AI and Microsoft partner on AI research as well.

While third-party rankings have yet to fully assess Mistral Large, the firm’s earlier Mistral Medium ranked 6th out of over 60 language models. With the latest release, Mistral AI appears positioned to challenge dominant players in the increasingly crowded AI space.

(Photo by Joshua Golde on Unsplash)

See also: Stability AI previews Stable Diffusion 3 text-to-image model

Want to learn more about AI and big data from industry leaders? Check out AI & Big Data Expo taking place in Amsterdam, California, and London. The comprehensive event is co-located with other leading events including BlockX, Digital Transformation Week, and Cyber Security & Cloud Expo.

Explore other upcoming enterprise technology events and webinars powered by TechForge here.

The post Mistral AI unveils LLM rivalling major players appeared first on AI News.

The tech giant is set to unveil new data centres in Madrid and has outlined its intention to construct additional centres in Aragon, catering to European companies and public entities. The increased European infrastructure aims to deliver Microsoft’s cloud services with heightened security, privacy, and data sovereignty measures, facilitating access to the company’s full suite of AI solutions for businesses and public administrations in the region.

According to an analysis by IDC, these new Microsoft data centres have the potential to contribute €8.4 billion to the national GDP and help to generate 69,000 jobs from 2026 to 2030.

The commitment to investment aligns with a collaborative agreement forged between the President of the Government, Pedro Sánchez, and Microsoft President Brad Smith. Under this collaboration, Microsoft and the Government of Spain will collaborate on various initiatives aimed at advancing responsible AI, enhancing citizen services, and bolstering national cybersecurity and resilience across Spanish companies, public bodies, and critical infrastructures.

This partnership operates within the framework of the National Strategy for Artificial Intelligence and the National Cybersecurity Strategy outlined by the Spanish government. It revolves around four key action points:

1. Extension of AI in public administration: Efforts will be directed towards modernising administrative processes and equipping officials with AI tools to boost efficiency. This includes deploying generative AI solutions and implementing AI training plans for officials.

2. Promotion of responsible AI: Microsoft will share its responsible AI design standards, along with implementation guides and best practices documentation, with the Spanish Agency for the Supervision of Artificial Intelligence (AESIA).

3. Strengthening national cybersecurity: Collaboration with the National Cryptological Center (CNI) aims to enhance early warning mechanisms and response to cybersecurity incidents in public administrations.

4. Improving cyber-resilience of companies: Microsoft will collaborate with the National Institute of Cybersecurity (INCIBE) to enhance the cybersecurity posture of Spanish companies, particularly SMEs, by providing access to threat intelligence and conducting joint outreach initiatives.

Microsoft’s increased investment underscores its commitment to advancing technological innovation in Spain while fostering a secure and responsible digital ecosystem.

(Photo by engin akyurt on Unsplash)

See also: Wipro and IBM collaborate to propel enterprise AI

Want to learn more about AI and big data from industry leaders? Check out AI & Big Data Expo taking place in Amsterdam, California, and London. The comprehensive event is co-located with other leading events including BlockX, Digital Transformation Week, and Cyber Security & Cloud Expo.

Explore other upcoming enterprise technology events and webinars powered by TechForge here.

The post Microsoft is quadrupling its AI and cloud investment in Spain appeared first on AI News.

Phi-2 builds upon the success of its predecessors, Phi-1 and Phi-1.5, by matching or surpassing models up to 25 times larger—thanks to innovations in model scaling and training data curation.

The compact size of Phi-2 makes it an ideal playground for researchers, facilitating exploration in mechanistic interpretability, safety improvements, and fine-tuning experimentation across various tasks.

Phi-2’s achievements are underpinned by two key aspects:

• Training data quality: Microsoft emphasises the critical role of training data quality in model performance. Phi-2 leverages “textbook-quality” data, focusing on synthetic datasets designed to impart common sense reasoning and general knowledge. The training corpus is augmented with carefully selected web data, filtered based on educational value and content quality.

• Innovative scaling techniques: Microsoft adopts innovative techniques to scale up Phi-2 from its predecessor, Phi-1.5. Knowledge transfer from the 1.3 billion parameter model accelerates training convergence, leading to a clear boost in benchmark scores.

Performance evaluation

Phi-2 has undergone rigorous evaluation across various benchmarks, including Big Bench Hard, commonsense reasoning, language understanding, math, and coding.

With only 2.7 billion parameters, Phi-2 outperforms larger models – including Mistral and Llama-2 – and matches or outperforms Google’s recently-announced Gemini Nano 2:

Beyond benchmarks, Phi-2 showcases its capabilities in real-world scenarios. Tests involving prompts commonly used in the research community reveal Phi-2’s prowess in solving physics problems and correcting student mistakes, showcasing its versatility beyond standard evaluations:

Phi-2 is a Transformer-based model with a next-word prediction objective, trained on 1.4 trillion tokens from synthetic and web datasets. The training process – conducted on 96 A100 GPUs over 14 days – focuses on maintaining a high level of safety and claims to surpass open-source models in terms of toxicity and bias.

With the announcement of Phi-2, Microsoft continues to push the boundaries of what smaller base language models can achieve.

(Image Credit: Microsoft)

See also: AI & Big Data Expo: Demystifying AI and seeing past the hype

Want to learn more about AI and big data from industry leaders? Check out AI & Big Data Expo taking place in Amsterdam, California, and London. The comprehensive event is co-located with Digital Transformation Week.

Explore other upcoming enterprise technology events and webinars powered by TechForge here.

The post Microsoft unveils 2.7B parameter language model Phi-2 appeared first on AI News.