logo
Mistral launches Europe's first AI reasoning model

Mistral launches Europe's first AI reasoning model

Tahawul Tech4 days ago

Mistral recently launched Europe's first AI reasoning model as it tries to keep pace with American and Chinese rivals within the sphere of AI development. The French startup has attempted to differentiate itself by championing its European roots as well as making some of its models open source in contrast to other proprietary offerings.
Mistral is considered Europe's best shot at having a home-grown AI competitor, but has lagged behind in terms of market share and revenue. Reasoning models use chain-of-thought techniques – a process that generates answers with intermediate reasoning abilities when solving complex problems. They could also be a promising path forward in advancing AI's capabilities as the traditional approach of building ever-bigger large language models by adding more data and computing power begins to hit limitations.
For Mistral, which was valued by venture capitalists at $6.2 billion, an industry shift away from 'scaling up' could give it a window to catch up against better capitalised rivals. Mistral is launching an open-sourced Magistral Small model and a more powerful version called Magistral Medium for business customers.
'The best human thinking isn't linear – it weaves through logic, insight, uncertainty, and discovery. Reasoning language models have enabled us to augment and delegate complex thinking and deep understanding to AI,' Mistral said.
American companies have mostly kept their most advanced models proprietary, though a handful, such as Meta, has released open-source models. In contrast, Chinese firms such as DeepSeek have taken the open-source path to demonstrate their technological capabilities.
Source: Reuters
Image Credit: Dado Ruvic

Orange background

Try Our AI Features

Explore what Daily8 AI can do for you:

Comments

No comments yet...

Related Articles

Meta's $14.8 billion Scale AI deal latest test of AI partnerships
Meta's $14.8 billion Scale AI deal latest test of AI partnerships

Khaleej Times

time7 hours ago

  • Khaleej Times

Meta's $14.8 billion Scale AI deal latest test of AI partnerships

Facebook owner Meta's $14.8 billion investment in Scale AI and hiring of the data-labeling startup's CEO will test how the Trump administration views so-called acquihire deals, which some have criticized as an attempt to evade regulatory scrutiny. The deal, announced on Thursday, was Meta's second-largest investment to date. It gives the owner of Facebook a 49% nonvoting stake in Scale AI, which uses gig workers to manually label data and includes among its customers Meta competitors Microsoft and ChatGPT creator OpenAI. Unlike an acquisition or a transaction that would give Meta a controlling stake, the deal does not require a review by U.S. antitrust regulators. However, they could probe the deal if they believe it was structured to avoid those requirements or harm competition. The deal appeared to be structured to avoid potential pitfalls, such as cutting off competitors' access to Scale's services or giving Meta an inside view into rivals' operations - though Reuters exclusively reported on Friday that Alphabet's Google has decided to sever ties with Scale in light of Meta's stake, and other customers are looking at taking a step back. In a statement, a Scale AI spokesperson said its business, which spans work with major companies and governments, remains strong, as it is committed to protecting customer data. The company declined to comment on specifics with Google. Alexandr Wang, Scale's 28-year-old CEO who is coming to Meta as part of the deal, will remain on Scale's board but will have appropriate restrictions placed around his access to information, two sources familiar with the move confirmed. Large tech companies likely perceive the regulatory environment for AI partnerships as easier to navigate under President Donald Trump than under former President Joe Biden, said William Kovacic, director of the competition law center at George Washington University. Trump's antitrust enforcers have said they do not want to regulate how AI develops, but have also displayed a suspicion of large tech platforms, he added. "That would lead me to think they will keep looking carefully at what the firms do. It does not necessarily dictate that they will intervene in a way that would discourage the relationships," Kovacic said. Federal Trade Commission probes into past "aquihire" deals appear to be at a standstill. Under the Biden administration, the FTC opened inquiries into Amazon's deal to hire top executives and researchers from AI startup Adept, and Microsoft's $650 million deal with Inflection AI. The latter allowed Microsoft to use Inflection's models and hire most of the startup's staff, including its co-founders. Amazon's deal closed without further action from the regulator, a source familiar with the matter confirmed. And, more than a year after its initial inquiry, the FTC has so far taken no enforcement action against Microsoft over Inflection, though a larger probe over practices at the software giant is ongoing. A spokesperson for the FTC declined to comment. David Olson, a professor who teaches antitrust law at Boston College Law School, said it was smart of Meta to take a minority nonvoting stake. "I think that does give them a lot of protection if someone comes after them," he said, adding that it was still possible that the FTC would want to review the agreement. The Meta deal has its skeptics. U.S. Senator Elizabeth Warren, a Democrat from Massachusetts who is probing AI partnerships involving Microsoft and Google, said Meta's investment should be scrutinized. 'Meta can call this deal whatever it wants - but if it violates federal law because it unlawfully squashes competition or makes it easier for Meta to illegally dominate, antitrust enforcers should investigate and block it," she said in a statement on Friday. While Meta faces its own monopoly lawsuit by the FTC, it remains to be seen whether the agency will have any questions about its Scale investment. The U.S. Department of Justice's antitrust division, led by former JD Vance adviser Gail Slater, recently started looking into whether Google's partnership with chatbot creator was designed to evade antitrust review, Bloomberg News reported. The DOJ is separately seeking to make Google give it advance notice of new AI investments as part of a proposal to curb the company's dominance in online search.

Unglamorous world of 'data infrastructure' driving hot tech M&A market in AI race
Unglamorous world of 'data infrastructure' driving hot tech M&A market in AI race

Khaleej Times

time7 hours ago

  • Khaleej Times

Unglamorous world of 'data infrastructure' driving hot tech M&A market in AI race

Weighed down by tariffs and geopolitical uncertainty, dealmaking has slowed to a crawl across most industries except one: the unglamorous world of data infrastructure. The companies that process the data used to build advanced AI models have become highly sought after targets for legacy tech companies like Meta, Salesforce, and ServiceNow in the scramble to stay competitive against the likes of OpenAI, Google and Anthropic. 'AI without data is like life without oxygen, it doesn't exist,' said Brian Marshall, global co-head of software investment banking at Citi. 'Because of that, data is having a zeitgeist moment right now driven by AI,' Marshall said. Tech deals are one of the few bright spots in an otherwise gloomy MA market, accounting for $421 billion of the $1.67 trillion in global deals announced in the first five months of the year, or about 25% of total MA, according to preliminary data compiled for Reuters by Dealogic. That's up from about 20% last year and 17% in 2023, the data shows. Of the tech deals, those involving AI software makers accounted for almost three quarters of the overall value, the data shows. Speed matters Goldman Sachs Managing Director Matthew Lucas, who focuses on M&A relating to all aspects of computing, said enterprise data, as it applies to use in AI, is the 'most dynamic area in software MA right now.' 'There's a very strong perception that speed matters a lot, getting there first matters a lot, and that tends to lend itself to doing MA,' Lucas said. The software firms that help companies manage their data on cloud-based systems are increasingly valuable commodities as the number of such potential targets is rapidly shrinking. Enterprise data infrastructure and analytics companies like Confluent, Collibra, Sigma Computing, Matillion, Dataiku, Fivetran, Boomi, and Qlik, could become targets for legacy tech providers in the near term, investment bankers say. The companies, they say, may help businesses integrate, analyze, and store information better. Executives for Boomi, Dataiku, Fivetran, and Qlik all said they weren't surprised by the attention. "Messy, siloed data has long undermined the attempts of enterprises to deliver on the transformative potential of analytics. Now, with the urgency to deploy effective AI, fixing it isn't just essential — it's existential," Florian Douetteau, co-founder and CEO of Dataiku, said in a statement. Requests for comment from Confluent, Collibra, Sigma Computing, and Matillion weren't immediately returned. Legacy tech buys in Several multibillion-dollar deals for data infrastructure companies have been struck or closed just in the last few weeks. Meta announced on Friday a $14.8 billion deal for a 49% stake in data-labeling company Scale AI. Salesforce announced plans last month to buy data integration company Informatica for $8 billion. Artificial intelligence is driving a once-in-a-generation makeover in tech that's forcing several of the largest social media platforms and software makers to buy companies that help AI-backed systems run smoothly. Worldwide, generative AI spending is expected to total $644 billion in 2025, an increase of 76.4% from 2024, according to a forecast by technology data provider Gartner. In early May, IT management provider ServiceNow said it was buying data catalogue platform which will allow ServiceNow to better understand the business context behind data, executives said when it was announced. The Salesforce acquisition of Informatica, announced late last month, will allow Salesforce to better analyze and assimilate scattered data from across its internal and external systems before feeding it into its in-house AI system, Einstein AI, executives said at the time. IBM closed on its acquisition of data management provider DataStax the very next day. IBM said the deal, announced in February, will allow it to manage and process unstructured data before feeding it to its AI platform. Bad AI advice Those deals highlight the strategic importance for legacy software players to own all aspects of data management, and MA is often the fastest way to achieve it. Instead of building complex data systems from scratch, they are acquiring specialists that can help organize, clean, and connect data from across their business. Would-be targets have sometimes become the hunters as was the case when Databricks, a leader in data processing and AI that was recently valued at $62 billion, announced plans last week to buy serverless database manager Neon for $1 billion. But dealmakers warned that companies can't just buy any kind of data and throw it into an AI system and expect good results. Air Canada was found liable in small claims court and forced to refund airfare last year after one of its AI chatbots gave a customer bad advice. Those types of errors can happen if the wrong kind of unfiltered data is imported into an AI engine, tech dealmakers say. 'A lot of companies have a huge amount of data, but I think they're learning that you can't just funnel every piece of data you have into an AI engine with no organization, and hope that it spits out the right answer,' said Brian Mangino, partner at Latham Watkins.

Saudi Arabia showcases potential unicorns at VivaTech Paris
Saudi Arabia showcases potential unicorns at VivaTech Paris

Zawya

timea day ago

  • Zawya

Saudi Arabia showcases potential unicorns at VivaTech Paris

PARIS - (GLOBE NEWSWIRE) -- Monsha'at, Saudi Arabia's Small and Medium Enterprises General Authority, led a group of five potential Saudi unicorns SMEs to participate in VivaTech 2025, Europe's leading technology and innovation exhibition. Its participation was part of Saudi Unicorns; a Saudi Vision 2030 initiative to help high-potential, fast-growing companies in promising sectors reach valuations of over $1 billion. Held in Paris from June 11-14, the four-day innovation-focused event featured 13,500 startups, 3,200 investors, and over 165,000 attendees. Now in its ninth year, the conference was a key opportunity for Monsha'at to expose high-potential Saudi startups to global best practices, industry insights, and strategic networks with leading global stakeholders, gain fresh perspectives, and acquire the skills and contacts they need to globally scale. It also allowed them to share Saudi Arabia's success story with leading tech investors, which aligns with the Saudi Unicorns initiative's commitment to driving growth and innovation in the Saudi startup space. 'Our participation at VivaTech 2025 was a tremendous opportunity to not only showcase several of the best young startups and concepts coming out of the Kingdom,' said Saud Alsabhan, Vice Governor of Entrepreneurship at Monsha'at. 'It was also a chance to allow some of our best tech entrepreneurs to share insights about the Kingdom's booming startup ecosystem with leading global investors and innovators.' The Monsha'at-led Saudi cohort included five startups: Telgani, a digital car rental platform with over 3 million users; Soum, a second-hand sales app with over 6 million downloads; Webook, a digital engagement platform that has processed over $500 million in event transactions; SiFi, which provides financial solutions to over 500 Saudi companies; and Rewaa, a retail tech company serving over 7,000 clients across the Kingdom. Together, these companies showcase the dynamism of Saudi Arabia's growing e-commerce, transportation, entertainment technology, fintech, and retail tech sectors. Their involvement comes at a time of strong momentum for the Kingdom's SME and startup ecosystems. In Q1 2025, Saudi Arabia's private sector recorded a 48% q-o-q increase in commercial registrations, reaching 1.68 million active registrations nationwide. High-growth sectors included e-commerce, which saw a 6% increase to 41,322 registered entities, and cloud-computing, which surged 33% during the same period. Monsha'at's presence at VivaTech 2025 underlines Saudi Arabia's commitment to fostering entrepreneurship, enabling international partnerships, and accelerating the growth of Saudi tech leaders, particularly SMEs.

DOWNLOAD THE APP

Get Started Now: Download the App

Ready to dive into the world of global news and events? Download our app today from your preferred app store and start exploring.
app-storeplay-store