“Calm After the Storm”: Breaking Down Recent AI Updates.

Let’s be clear: for most of us—especially those who’ve been following the latest headlines—when it comes to AI and its updates, the first question that comes to mind is, “What is really going on here?”

After whirlwind years of breakthroughs, investments, and nonstop hype, the generative AI market seems to be settling into a new phase—one of strategic expansion and measured growth. While the initial gold rush may be over, AI development companies are far from slowing down. Instead, they’re scaling up, refining their models, and focusing on long-term impact.

These days, it’s easy to get lost in all the DeepSeek this, Mistral that. So today, we’re inviting you to dive a little deeper into the latest updates in the AI industry—and, hopefully, make it all a bit less confusing.

Let’s clear the air, shall we?

The Blooming Market

Despite economic fluctuations, funding for AI startups and enterprise solutions continues to soar.?

According to recent market reports, 微软 leads the generative AI space with a 39% market share, followed by Amazon Web Services (AWS) (19%) and 谷歌 (15%). OpenAI and Anthropic continue to make waves, securing significant funding, while companies like 阿里巴巴集团 and 百度 remain strong contenders in the global AI race.

The financial side of AI is just as impressive. The industry is experiencing remarkable growth, with startups attracting substantial investments and achieving significant valuations. According to Forbes’ 2024 AI 50 list, the top AI startups have collectively raised $34.7 billion, up from $27.2 billion the previous year.

?? Forbes 2024 AI 50 List - Top Artificial Intelligence Startups?

The competition is fierce, but so is the demand. Industries from healthcare to finance are integrating AI at record speeds. So, the question of? “whether the technology is here to stay?” is no longer relevant. What is more accurate,

How does business encourage technology to expand its influence?

Businesses Support AI Thriving

Safe to say, 2024 has undoubtedly become the year of massive AI implementation across industries and businesses. With a notable 17% surge in adoption and a dramatic 2x increase in usage, AI has shifted from a “nice-to-have” to a competitive necessity.

麦肯锡 research highlights that just a few business functions account for ~75% of generative AI’s total economic impact. The biggest winners? Sales, marketing, software engineering, and customer operations, each generating hundreds of billions in value.

We’re now far enough into the gen AI era to see patterns among companies that are capturing value. One significant difference is that these companies focus as much on driving adoption and scaling as they do on the up-front technology development. - Bryce Hall

Beyond the buzz, AI’s real appeal lies in its practical benefits for businesses. Leaders are no longer just exploring AI—they’re expecting it to solve real-world challenges and unlock measurable value. From driving revenue to boosting efficiency, the expectations are clear and ambitious.

What businesses are actually getting from AI? Here are a couple of impressive examples:

Retail

• Walmart is using generative AI to boost employee productivity and customer experience. In a pilot project, an AI chatbot automated negotiations with suppliers, closing deals with 64% of them. This resulted in 1.5% cost savings and 35 extra days in payment terms. 83% of suppliers had a positive experience, highlighting the AI’s efficiency.

Banking

• JPMorgan Chase implemented an AI coding assistant that increased the productivity of its software engineers by up to 20%. This tool allows engineers to focus more on high-value projects, particularly in artificial intelligence and data. With a $17 billion tech budget for 2025 and a tech workforce of 63,000, JPMorgan plans to leverage AI across approximately 450 potential use cases, potentially increasing to 1,000 next year.?

Healthcare

• NIB, an Australian health insurer, has saved $22 million since 2021 by implementing AI through its digital assistant, Nibby, which handles simpler customer inquiries, reducing the need for human digital support by 60% and phone calls by 15%.

Manufacturing and energy

• ACG Capsules, a global leader in the manufacturing of empty hard capsules and pharmaceutical machinery, implemented an AI copilot that reduced repair times and technical onboarding efforts by nearly 40%. The technology was developed and deployed in just five weeks.

New Players Join The Party

After the era of AI giants like OpenAI, Google, and Microsoft took the lead on the AI scene, competition is now opening doors for fresh contenders.

DeepSeek Shakes Up Stock Market

This story begins back in 2023, when a Chinese AI startup, DeepSeek, launched—but it took a whole new twist two years later.

In early 2025, DeepSeek introduced its R1 model, which matched the performance of established AI systems like OpenAI’s ChatGPT at a fraction of the cost. Not only did it outperform existing models, but it also offered scalable, cost-effective solutions.

And then the drama began. DeepSeek’s AI model, a direct competitor to ChatGPT, quickly surged to the top of the App Store and caused significant market disruption. On January 27th, Nvidia’s stock dropped by 17%, wiping out nearly $600 billion from its market value. CNBC reported that this marked the largest single-day decline in U.S. history.

How did this happen? In 2021, CEO Liang Wenfeng set the stage for DeepSeek’s breakthrough by purchasing thousands of Nvidia GPUs, paving the way for the creation of a revolutionary AI model. The team managed to slash R1’s training costs by 95%, bringing the total to just $5.6 million—far lower than OpenAI’s. Instead of building from the ground up, they strategically leveraged open-source models, starting with Meta’s Llama, to accelerate development.

Naturally, this story didn’t end there. From facing a potential ban on U.S. government devices to dealing with allegations of exposed user chat histories, DeepSeek’s R1 continues to make waves. Meanwhile, Microsoft is now working to make the R1 model available on Azure AI and GitHub.

Something tells us this saga is far from over.

?? DeepSeek: all the news about the startup that’s shaking up AI stocks

Mistral AI’s? “Le Chat” Enters The Chat

Moving on from the drama to une histoire incroyable. Yes, there’s a reason we couldn’t resist switching to a little French—and it’s Mistral AI, one of the most impressive recent stories in the AI world.

Mistral AI, a Paris-based startup founded by ex-Google and Meta researchers, is quickly rising to prominence. In 2024, the company raised $644 million, valuing it at $6.2 billion. Known for its sparse Mixture of Experts architecture, Mistral delivers highly efficient AI models that rival the performance of larger competitors like OpenAI and Anthropic.

With its rapid growth, Mistral is helping position Europe as a key player in the global AI race, challenging U.S. dominance in the space—and recent news proves it!

In late February, just two weeks after unveiling Le Chat, Mistral’s AI assistant, the app hit a remarkable milestone—one million downloads. It quickly soared to the top, claiming the #1 spot for free downloads on the iOS App Store in France. But there’s even more.

By February 2025, Mistral AI had raised nearly €1 billion in capital—about $1.04 billion—solidifying its position as one of Europe’s most well-funded AI startups.

We’ll definitely be keeping our eyes on this one to see what’s next!

?? What is Mistral AI? Everything to know about the OpenAI competitor

AI Giants Is Here To Stay As Well

In 2025, AI giants are clearly following a common pattern: going multimodal, enhancing reasoning capabilities, and focusing on personalization.?

One of the recent McKinsey reports illustrates a sharp evolution from 2022-23 models, which were mostly text-only and struggled with complex conversations, to today’s systems that handle text, audio, and images.?

Companies like OpenAI, Google, and Anthropic have prioritized long-context coherence and advanced reasoning for nuanced analysis. Real-time data integration (Google) and experimental tool usage (Anthropic) hint at a race toward more autonomous, versatile AI systems. The trend? Faster, smarter, more user-adaptive models.

Another trend that goes far beyond GenAI models is companies focusing on building in-house chips. According to Reuters, Meta is reportedly testing its own AI chips designed specifically to handle AI workloads, mainly for training AI systems. Similarly, OpenAI is said to be getting closer to launching its own in-house chip.

Meanwhile, NVIDIA is gearing up to unveil its next AI chip—Blackwell Ultra—later this March, with full availability expected in the second half of next year. The upgrade promises new networking capabilities, memory enhancements, and processors.

“As demand for generative AI services continues to grow, it’s evident that chips will be the next big battleground for AI supremacy.” - The Verge

All of this brings us to the question…

What Comes Next?

Now that we’ve got a clearer picture of recent AI developments, let’s take a look at a few trends that are likely just around the corner.

According to McKinsey, Gen AI innovation today is marked by four clear tendencies:

1. Agentic AI: AI systems are increasingly capable of autonomously reasoning and executing complex tasks, moving from supporting roles to fully handling workflows such as customer service and operations.
2. Multimodality: AI models are now integrating and processing text, audio, and video together, enabling more human-like, emotionally nuanced interactions and advanced media generation.
3. Hardware innovation: Advances in specialized hardware like GPUs, TPUs, and edge devices are boosting AI model performance, scalability, and enabling real-time, high-demand applications.
4. Transparency: AI models are steadily improving transparency and explainability, crucial for reducing risk, ensuring compliance, and building trust in enterprise use.

A recent study from MIT Technology Review went beyond GenAI, also predicting the main sectors AI will probably conquer next.

• Generative virtual worlds (a.k.a. video games): Advances in generative AI are transforming video games, with models like Google DeepMind’s Genie creating playable virtual worlds from images. These technologies also have applications in training robots and developing spatial intelligence.
• Science: AI is accelerating scientific discoveries, exemplified by DeepMind’s AlphaFold solving protein folding, and the pursuit of AI-driven advancements in materials science and other fields. AI tools are becoming increasingly useful in scientific research, potentially acting as virtual researchers in the future.
• National security: AI is being rapidly adopted in national security, with military programs like the US Replicator initiative investing in AI-driven defense tools. Major defense-tech companies are capitalizing on this demand, with AI companies like OpenAI now partnering with the military for projects such as drone defense.

?? What’s next for AI in 2025

As we stand at this turning point, it’s clear that AI is evolving at an astonishing pace—becoming smarter, more autonomous, and increasingly embedded. Much has already changed, yet it feels like we’re only at the very beginning of what AI will offer next.