hodler

4 Mar 2026, 19:20

Two-thirds of European firms use AI, but only 25% actually invest in the growing technology

The adoption of AI within European businesses is on a steady rise; however, the numbers show that most companies aren’t actually paying for it. In a research published by the European Central Bank (ECB), the use of AI has become widespread across continents, but actual investments in the technology have not produced the same results due to companies relying on free tools rather than searching for enterprise solutions. The ECB’s post was compiled after the bank’s survey on Access to Finance Enterprises, which was carried out between the second and fourth quarters of 2025. Why are companies not investing despite widespread use? A major reason for the divide between usage and investment levels lies in the issue of accessibility. Most firms do not see a reason to invest in AI infrastructure to deploy the technology, because accessible tools like ChatGPT, Claude , open-source AI models, and specific browser extensions have drastically dropped the barrier to entry. With these tools, companies can equip their entire workforce with AI capabilities without having to dip into company funds and without requiring custom solutions. According to the ECB, 90% of businesses with 250 or more employees make use of AI, compared to companies with 10 employees or fewer. On the other hand, investment in AI capabilities drops to around one in every four companies across the board. This greatly impacts the effects of AI on the economy. As the technology keeps developing and adoption increases, the capital expenditure isn’t growing at the same rate, suggesting that companies would rather experiment with AI freely rather than commit funds to it. Are firms replacing workers with AI? According to the ECB’s findings , companies using AI are not looking to replace workers, but are 4% more likely to hire additional staff than firms that do not. Additionally, businesses that invest in AI are 2% more likely to grow their workforce. This pattern occurs more often in smaller companies, while larger firms are not affected by AI adoption, suggesting that AI is more of a tool in smaller companies than an employee replacement. This is because these firms primarily use AI for research, development, and innovation applications to increase productivity and not to automate existing tasks. AI has taken a different route from past adoption predictions The ECB’s findings do not match the results from earlier research projects, such as the survey conducted by Germany’s Ifo Institute. The institute concluded from its survey that over 25% of German companies believed that AI would reduce the workforce within five years. Additionally, major companies in the US, such as Amazon, have linked thousands of job cuts to AI reasons. This difference can be attributed to timing and geography. The ECB’s research was conducted around what’s happening now and over the next year in Europe, where AI adoption varies differently when compared to the United States . For example, European companies have stricter rules when approaching AI investment and workforce structure. Another difference is the scale of investment in AI. According to Lebastard and Sonderman, the extent and timing of AI adoption differ between the US and Europe, pointing out how AI has had little effect on how Europeans conduct their business, and functions more like a support than a core aspect of their production. Lastly, in a paper published in January by the European Investment Bank , most firms that adopted AI boosted productivity by 4% through capital investment, and not through job cuts. The productivity boost often occurred in medium and large-sized organizations, with AI-adopting firms paying higher wages and incurring more innovative costs. If you're reading this, you’re already ahead. Stay there with our newsletter .

4 Mar 2026, 18:50

Perplexity signs a multi-year deal to run AI workloads on CoreWeave's infrastructure

Perplexity, the AI search company, signed a multi-year deal to run its AI workloads on CoreWeave’s cloud platform, and investors took notice, pushing CRWV shares up roughly 5.7% in pre-market trading. The deal puts Perplexity on NVIDIA GB200 NVL72-powered clusters through CoreWeave’s infrastructure. Those clusters will carry the load for Perplexity’s fast-growing AI products, along with its Sonar and Search API services. CoreWeave is also bringing Perplexity Enterprise Max into its own offices. Staff will use it to search the web, pull from internal knowledge bases, run multi-step research, look through data, and tap into advanced AI models, all from a single place. Perplexity has already started running workloads through CoreWeave’s Kubernetes service as part of its first deployment phase. It is also using W&B Models to train, fine-tune, and manage its models from early testing through to live production. The move fits Perplexity’s wider strategy of spreading its infrastructure across more than one cloud provider, while adding to CoreWeave’s growing list of AI clients running at production scale. Max Hjelm, CoreWeave’s SVP of Revenue, sai d pr oduction AI demands more than raw computing power. “AI applications running in production require more than just access to raw infrastructure; they require best-in-class performance and reliability as well as a cloud platform designed end-to-end for AI that simplifies compute operations,” he said. Perplexity’s Chief Business Officer Dmitry Shevelenko called CoreWeav e an “essential partner” for where the company is headed. div]:bg-bg-000/50 [&_pre>div]:border-0.5 [&_pre>div]:border-border-400 [&_.ignore-pre-bg>div]:bg-transparent [&_.standard-markdown_:is(p,blockquote,h1,h2,h3,h4,h5,h6)]:pl-2 [&_.standard-markdown_:is(p,blockquote,ul,ol,h1,h2,h3,h4,h5,h6)]:pr-8 [&_.progressive-markdown_:is(p,blockquote,h1,h2,h3,h4,h5,h6)]:pl-2 [&_.progressive-markdown_:is(p,blockquote,ul,ol,h1,h2,h3,h4,h5,h6)]:pr-8"> _*]:min-w-0 gap-3 standard-markdown"> Fresh off an 8% drop Shares fell 8% in extended tradin g on Th ursday after an earnings report showed widening losses and a weaker outlook than Wall Street had expected, despite strong revenue. The company’s contracted revenue backlog came in at $66.8 billion, which points to strong long-term demand, though concerns about execution and heavy reliance on a handful of customers have kept some investors cautious. Looking ahead, CoreWeave is planning to spend between $30 billion and $35 billion on capital expenditures in 2026, a sharp jump from $10.31 billion in 2025. It wants to hit more than 1.7 gigawatts of active power by year-end, ahead of the analyst consensus sitting at 1.59 gigawatts, and grow beyond five gigawatts past its contracted footprint by 2030. div]:bg-bg-000/50 [&_pre>div]:border-0.5 [&_pre>div]:border-border-400 [&_.ignore-pre-bg>div]:bg-transparent [&_.standard-markdown_:is(p,blockquote,h1,h2,h3,h4,h5,h6)]:pl-2 [&_.standard-markdown_:is(p,blockquote,ul,ol,h1,h2,h3,h4,h5,h6)]:pr-8 [&_.progressive-markdown_:is(p,blockquote,h1,h2,h3,h4,h5,h6)]:pl-2 [&_.progressive-markdown_:is(p,blockquote,ul,ol,h1,h2,h3,h4,h5,h6)]:pr-8"> _*]:min-w-0 gap-3 standard-markdown">A well-timed announcement before investor conferences The partnership gives CoreWeave a high-profile new customer outside its Microsoft/OpenAI concentration problem, fresh ammunition for diversification, and a premarket stock bump, all before the investor conference . Co-Founder and Chief Development Officer Brannin McBee will speak at the Morgan Stanley TMT Conference in San Francisco on Wednesday, March 4, 2026, starting at 4:05 p.m. Eastern. Vice President Nick Robbins will follow at the Cantor Global Technology Conference in New York on Tuesday, March 10, 2026, at 2:30 p.m. Eastern. As previously reported by Cryptopolitan , Nvidia has put $2 billion into CoreWeave, picking up Class A shares at $87.20 each. CEO Mike Intrator said th e money will help the company “accelerate our build” and spread its customer base. “This will lead to continued diversification,” he said. CoreWeave makes its money by renting out GPU-heavy computing capacity from its data centers, the kind of muscle companies need to train AI models and keep them running. That puts it in a growing class of cloud providers buil t fo r one thing: powering AI. Want your project in front of crypto’s top minds? Feature it in our next industry report, where data meets impact.

4 Mar 2026, 18:20

Denmark Defense Spending: Resilient Fiscal Strength Meets Soaring Security Demands – Nordea Analysis

BitcoinWorld Denmark Defense Spending: Resilient Fiscal Strength Meets Soaring Security Demands – Nordea Analysis COPENHAGEN, March 2025 – Denmark’s renowned fiscal discipline now confronts a transformative challenge as escalating security demands test the nation’s economic resilience. According to a comprehensive analysis by Nordea, Scandinavia’s largest financial services group, the Danish state must navigate unprecedented defense expenditure increases while maintaining its AAA credit rating and welfare model. This pivotal moment represents a critical stress test for one of Europe’s most stable economies. Denmark Defense Spending in Historical Context Denmark has consistently maintained defense expenditures below the NATO target of 2% of GDP for decades. However, the geopolitical landscape shifted dramatically following Russia’s invasion of Ukraine in 2022. Consequently, the Danish parliament approved a historic defense agreement in 2023, committing to reach the 2% target by 2030. This decision marked a fundamental reorientation of Danish security policy. Meanwhile, Nordea economists emphasize that Denmark enters this period from a position of exceptional fiscal strength. The Danish government debt-to-GDP ratio stands at approximately 30%, significantly below the Eurozone average of 90%. This robust balance sheet provides crucial maneuvering room for increased investments. The Fiscal Architecture Supporting Danish Resilience Denmark’s economic framework features several unique characteristics that bolster its capacity to absorb spending shocks. The country operates a flexible labor market with high participation rates, supported by its famous “flexicurity” model. Additionally, Denmark maintains substantial foreign exchange reserves and a current account surplus. Nordea’s analysis highlights three core pillars of Danish fiscal strength: Structural Budget Surpluses: Denmark has recorded budget surpluses in most years since 2015, creating fiscal buffers. Low Debt Servicing Costs: With AAA-rated sovereign debt, Denmark borrows at historically low interest rates. Counter-Cyclical Policies: Automatic stabilizers and discretionary measures provide economic shock absorption. These factors collectively enable strategic investments without jeopardizing long-term sustainability. Nevertheless, defense spending increases arrive alongside other pressures, including green transition costs and demographic aging. Nordea’s Quantitative Assessment of Defense Impact Nordea economists have modeled multiple scenarios for defense expenditure growth through 2030. Their baseline projection anticipates defense spending rising from 1.4% of GDP in 2023 to 2.0% by 2030, representing a cumulative increase of approximately 85 billion Danish kroner. Importantly, this expansion occurs alongside existing commitments to healthcare, education, and climate initiatives. The analysis suggests several potential economic effects: Economic Indicator 2023 Baseline 2030 Projection Change Defense Spending (% of GDP) 1.4% 2.0% +0.6% Government Debt (% of GDP) 30.2% 32.8% +2.6% Budget Balance (% of GDP) +0.8% +0.2% -0.6% Military Personnel 20,000 25,000 +25% These projections assume moderate economic growth and no major external shocks. Significantly, the modeling indicates that Denmark can maintain its fiscal surplus tradition despite increased spending, though margins become narrower. Comparative Analysis with Nordic Neighbors Denmark’s defense spending trajectory aligns with broader Nordic security cooperation. Sweden and Finland, following their NATO accessions, have announced even more substantial defense budget increases. Norway, with its sovereign wealth fund, faces different fiscal constraints. Meanwhile, Nordea’s regional analysis reveals divergent approaches to financing security enhancements. Sweden plans temporary tax increases, while Finland utilizes borrowing within EU fiscal rules. Denmark’s strategy relies primarily on economic growth and expenditure reprioritization. This comparative perspective underscores Denmark’s relatively conservative fiscal approach, even during a period of strategic transformation. The Industrial and Technological Multiplier Effect Beyond direct budgetary impacts, increased defense spending generates significant economic ripple effects. Danish defense contractors like Terma and Systematic stand to benefit from procurement programs. Furthermore, research and development in cybersecurity, surveillance, and naval technology receives substantial funding boosts. Nordea analysts note that defense investments often catalyze civilian technological innovation, creating positive spillovers across the economy. However, they caution against overestimating these effects, as defense manufacturing represents a relatively small sector within Denmark’s service-dominated economy. The primary economic challenge remains balancing competing priorities within finite fiscal resources. Long-Term Sustainability and Intergenerational Equity Sustained defense spending increases raise important questions about intergenerational fairness. Current investments primarily benefit future security, yet financing occurs through present taxation or borrowing. Nordea’s intertemporal analysis examines whether Denmark’s fiscal framework adequately addresses this temporal mismatch. The Danish government’s long-term fiscal sustainability report, published annually, now incorporates enhanced security spending scenarios. These projections help policymakers evaluate trade-offs between defense, welfare, and debt accumulation. Crucially, Denmark’s strong institutions and transparent budgeting processes facilitate informed democratic deliberation about these choices. Conclusion Denmark’s journey toward meeting NATO defense spending targets unfolds from a position of exceptional fiscal strength, according to Nordea’s comprehensive analysis. The nation’s low debt, consistent surpluses, and robust institutions provide a solid foundation for increased security investments. However, challenges emerge from competing priorities, including climate transition and demographic pressures. Ultimately, Denmark’s defense spending decisions will test the flexibility of its economic model while reinforcing its commitment to collective European security. The coming years will demonstrate whether Danish fiscal resilience can transform security necessities into sustainable strategic advantages. FAQs Q1: What percentage of GDP does Denmark currently spend on defense? Denmark’s defense spending reached approximately 1.7% of GDP in 2024, according to NATO estimates, with plans to achieve the 2% target by 2030 through gradual annual increases. Q2: How does Denmark’s defense burden compare to other NATO countries? Denmark traditionally spent below the NATO average but now aligns with European trends following Russia’s invasion of Ukraine. Several allies, including Poland and the Baltic states, exceed 2.5% of GDP, while major economies like Germany approach 2%. Q3: What are the main areas of increased Danish defense investment? Primary investment areas include naval capabilities (frigates and patrol vessels), air defense systems, cybersecurity infrastructure, and increased military personnel. The 2023 defense agreement specifically prioritizes Arctic surveillance and Baltic Sea security. Q4: How does Nordea assess the impact on Denmark’s AAA credit rating? Nordea analysts believe Denmark’s rating remains secure due to low initial debt, strong institutions, and gradual spending implementation. However, they note that simultaneous pressure from multiple spending areas could eventually test rating agencies’ assessments. Q5: What economic sectors benefit most from increased defense spending? Defense manufacturing, cybersecurity services, and specialized technology sectors experience direct benefits. Indirectly, construction, logistics, and professional services also gain from associated infrastructure and support contracts. This post Denmark Defense Spending: Resilient Fiscal Strength Meets Soaring Security Demands – Nordea Analysis first appeared on BitcoinWorld .

4 Mar 2026, 16:38

Apple iPhone Hacking Kit Used By Spies, Crypto Scams Could Have US Intelligence Origins

Researchers said a sophisticated exploit kit with 23 iOS vulnerabilities is being used by espionage and cybercrime campaigns.

4 Mar 2026, 16:33

Coinbase, Galaxy, bitcoin miners surge after Trump urges passage of stalled crypto bill

More on Coinbase, Galaxy Digital, etc. Coinbase Global, Inc. (COIN) Presents at Morgan Stanley Technology, Media & Telecom Conference 2026 Transcript Riot Platforms: A Hold At Best Riot Platforms, Inc. (RIOT) Q4 2025 Earnings Call Transcript Trump meets Coinbase CEO, criticizes banks over stalled crypto bill - Politico Galaxy to voluntarily delist from the TSX, maintain Nasdaq listing

4 Mar 2026, 15:50

Crowdsourced AI: One Startup’s Revolutionary Pitch to Crush Unreliable Chatbot Hallucinations

BitcoinWorld Crowdsourced AI: One Startup’s Revolutionary Pitch to Crush Unreliable Chatbot Hallucinations In Boston’s competitive tech landscape, a persistent problem plagues enterprises eager to harness artificial intelligence: unreliable answers. Frustrated by costly contracts and AI hallucinations, one CEO’s quest for accuracy sparked a novel solution—crowdsourcing the chatbots themselves. This innovative approach, emerging from Buyers Edge Platform’s incubation, represents a significant shift in how businesses might interact with large language models moving forward. Crowdsourced AI Emerges from Enterprise Frustration John Davie, CEO of hospitality procurement enterprise Buyers Edge Platform, initially embraced the AI wave with optimism. Consequently, he encouraged widespread experimentation among employees. However, this enthusiasm quickly collided with harsh realities. “We had a wake-up call,” Davie explained to Bitcoin World. “We learned that using various AI tools could mean training models on our company information.” This presented a clear security risk, potentially advantaging competitors. Furthermore, exploring secure enterprise options revealed another layer of issues. Davie discovered expensive, long-term contracts for LLMs that still produced inaccurate information and frequent hallucinations. The situation created internal tension. “We hated having to decide which employees deserved AI,” he stated. More critically, employees reported biased or flatly incorrect answers making their way into official presentations, undermining trust and productivity. The Technical Blueprint for Multi-Model Consensus Faced with these challenges, Davie tasked his chief technology officer with building a superior system. The result was CollectivIQ, a Boston-based spinout. Technically, the platform operates by querying several leading large language models simultaneously. This includes models from OpenAI (ChatGPT), Anthropic (Claude), Google (Gemini), and xAI (Grok), among up to ten others. The software’s core innovation lies in its consensus mechanism. It analyzes responses for overlapping and differing information, subsequently fusing them to generate a single, more accurate answer. This method leverages the collective strength of multiple AI systems to mitigate the weaknesses of any single one. For enterprise privacy, all prompt data is encrypted and deleted after use, addressing a primary concern for business adoption. Addressing the Hallucination Epidemic in Enterprise AI The issue of AI “hallucinations”—where models generate plausible but incorrect or fabricated information—has become a major barrier to professional adoption. A 2024 Stanford Institute for Human-Centered AI study highlighted that even advanced models exhibit hallucination rates between 15-20% in complex query scenarios. For businesses, this unreliability translates into real risk, from flawed market analyses to incorrect compliance information. CollectivIQ’s multi-model approach directly targets this problem. By cross-referencing answers, the system identifies outliers and inconsistencies that often signal hallucinations. This process is analogous to seeking multiple expert opinions before reaching a conclusion, thereby increasing confidence in the final output. The company’s usage-based pricing model also contrasts sharply with the industry norm of hefty upfront commitments, offering financial flexibility. Market Impact and the Evolving AI Landscape The launch of CollectivIQ arrives during a pivotal moment for enterprise AI. Following initial excitement, many companies now face a “trough of disillusionment” regarding implementation challenges. High costs, data security fears, and output reliability are causing hesitation. Davie noted that conversations with Buyers Edge Platform’s customers revealed widespread shared confusion, prompting the decision to publicly release the internally developed tool. This crowdsourcing model could influence how AI ecosystems evolve. Instead of a winner-take-all market dominated by one or two LLM providers, a future might emerge where applications leverage multiple specialized models. CollectivIQ builds its service using official enterprise APIs, paying token costs directly to model providers and passing on a consolidated, value-based charge to its customers. From Internal Tool to Public Venture After a strong internal rollout in early 2026, CollectivIQ is now seeking its place in the crowded enterprise AI market. Fully funded by Davie initially, the startup plans to seek outside capital later this year. For Davie, building CollectivIQ marks a return to startup scrappiness nearly three decades after launching his main company. “It’s fun and exciting,” he reflected. “I go sit hand in hand with the software developers… that’s how I got my main company.” The company’s success will likely depend on its ability to demonstrably reduce error rates and provide clear ROI compared to single-model subscriptions. Its approach also raises interesting questions about the future of AI benchmarking, potentially shifting focus from individual model performance to the effectiveness of ensemble methods. Conclusion CollectivIQ’s pitch for crowdsourced AI represents a pragmatic response to the reliability crisis facing enterprise artificial intelligence. By leveraging multiple LLMs to achieve consensus, the Boston startup offers a novel path toward more accurate, trustworthy business intelligence. As companies grow increasingly wary of AI hallucinations and data privacy, solutions that prioritize verification and flexibility may define the next phase of corporate AI adoption. The journey from internal frustration to public innovation underscores how hands-on experience continues to drive meaningful technological advancement. FAQs Q1: What is crowdsourced AI? Crowdsourced AI refers to a system that queries multiple artificial intelligence models simultaneously, comparing and synthesizing their responses to produce a single, more reliable answer. This approach aims to reduce errors and hallucinations common in single-model outputs. Q2: How does CollectivIQ ensure data privacy for enterprises? The company states that all data involved with user prompts is encrypted during processing and permanently deleted after use. This enterprise-grade privacy model is designed to prevent sensitive company information from being used to train public AI models. Q3: What problem does multi-model AI consensus solve? It primarily addresses the issue of AI hallucinations—incorrect or fabricated information generated by language models. By cross-referencing answers from multiple sources, the system can identify and filter out inconsistent or unreliable data points. Q4: How is CollectivIQ’s pricing model different? Unlike typical enterprise AI contracts that require expensive long-term commitments, CollectivIQ uses a pay-by-usage model. Customers incur costs based on their actual consumption of AI queries, rather than paying large upfront fees for seat licenses or compute credits. Q5: Which AI models does CollectivIQ currently integrate? The platform queries several leading large language models, including OpenAI’s ChatGPT, Anthropic’s Claude, Google’s Gemini, and xAI’s Grok. The system can pull from up to ten different models simultaneously to generate its fused answers. This post Crowdsourced AI: One Startup’s Revolutionary Pitch to Crush Unreliable Chatbot Hallucinations first appeared on BitcoinWorld .