It’s that time of year again, when those of us in the northern hemisphere pack our sunscreen and get ready to venture to hotter climates in search of some much-needed Vitamin D.

Every year, I book a vacation, and every year I get stressed as the big day gets closer, usually forgetting to pack something essential, like a charger for my Nintendo Switch 2, or dare I say it, my passport.

Enterprises may want to start thinking of large language models (LLMs) as ensemble casts that can combine knowledge and reasoning to complete tasks, according to Japanese AI lab Sakana AI.

Sakana AI in a research paper outlined a method called Multi-LLM AB-MCTS (Adaptive Branching Monte Carlo Tree Search) that uses a collection of LLMs to cooperate, perform trial-and-error and leverage strengths to solve complex problems.

In a post, Sakana AI said:

“Frontier AI models like ChatGPT, Gemini, Grok, and DeepSeek are evolving at a breathtaking pace amidst fierce competition. However, no matter how advanced they become, each model retains its own individuality stemming from its unique training data and methods. We see these biases and varied aptitudes not as limitations, but as precious resources for creating collective intelligence. Just as a dream team of diverse human experts tackles complex problems, AIs should also collaborate by bringing their unique strengths to the table.”

Sakana AI said AB-MCTS is a method for inference-time scaling to enable frontier AIs to cooperate and revisit problems and solutions. Sakana AI released the algorithm as an open source framework called TreeQuest, which has a flexible API that allows users to use AB-MCTS for tasks with multiple LLMs and custom scoring.

What’s interesting is that Sakana AI gets out of that zero-sum LLM argument. The companies behind LLM training would like you to think there’s one model to rule them all. And you’d do the same if you were spending so much on training models and wanted to lock in customers for scale and returns.

Sakana AI’s deceptively simple solution can only come from a company that’s not trying to play LLM leapfrog every few minutes. The power of AI is in the ability to maximize the potential of each LLM. Sakana AI said:

“We saw examples where problems that were unsolvable by any single LLM were solved by combining multiple LLMs. This went beyond simply assigning the best LLM to each problem. In (an) example, even though the solution initially generated by o4-mini was incorrect, DeepSeek-R1-0528 and Gemini-2.5-Pro were able to use it as a hint to arrive at the correct solution in the next step. This demonstrates that Multi-LLM AB-MCTS can flexibly combine frontier models to solve previously unsolvable problems, pushing the limits of what is achievable by using LLMs as a collective intelligence.”

A few thoughts:

• Sakana AI’s research and move to emphasize collective intelligence over on LLM and stack is critical to enterprises that need to create architectures that don’t lock them into one provider.
• AB-MCTS could play into what agentic AI needs to become to be effective and complement emerging standards such as Model Context Protocol (MCP) and Agent2Agent.
• If combining multiple models to solve problems becomes frictionless, the costs will plunge. Will you need to pay up for OpenAI when you can leverage LLMs like DeepSeek combined with Gemini and a few others? 
• Enterprises may want to start thinking about how to build decision engines instead of an overall AI stack. 
• We could see a scenario where a collective of LLMs achieves superintelligence before any one model or provider. If that scenario plays out, can LLM giants maintain valuations?
• The value in AI may not be in the infrastructure or foundational models in the long run, but the architecture and approaches.

More:

Mark Zuckerberg-led Meta Platforms (META) has proved its critics wrong as its shares have recently climbed to new heights, largely thanks to its artificial intelligence-driven strategy. Central to this AI strategy is “Superintelligence,” a long-term vision Zuckerberg has for creating AI systems that exceed human-level intelligence across many domains.

And although Zuckerberg burned shareholders before with the metaverse, his last passion project, Superintelligence feels different. Unlike the metaverse, AI is a megatrend that is already revolutionizing daily life. And Meta, with its arsenal of popular social media platforms like Instagram, WhatsApp, and Facebook, is betting big on AI to drive growth in the coming years. Meta is hiring big to staff this revolution, with Scale AI founder Alexandr Wang tasked with heading the new Superintelligence unit at Meta.

The market seems to be convinced this time, with Meta stock already up about 23% on a YTD basis.

Can Meta sustain this rally? I believe so, and here is why.

Meta has been doubling down on its AI ambitions, both by making significant financial commitments and by attracting top talent from rival firms. To that end, the company has reportedly extended compensation offers ranging from $50 million to $100 million to lure engineers away from OpenAI and Anthropic. It also made a $14.3 billion investment for a 49% stake in Scale AI, a startup recognized for its industry-leading data labeling capabilities. This investment positions Meta advantageously when it comes to securing high-quality training datasets.

With such resources in place, Mark Zuckerberg is equipping Meta’s AI models to be not just competitive, but potentially market-leading.

Meta’s powerful cash generation is giving it the flexibility to aggressively invest in AI infrastructure. The company has earmarked $60 billion to $72 billion for capital spending in 2025, much of which will be spent on building and upgrading data centers. This rapid pace of investment demonstrates Meta’s conviction that long-term value can be realized by investing in innovation-driven scale.

Story Continues

Investors should also be watching V-JEPA 2, the second iteration of Meta’s Video Joint Embedding Predictive Architecture. V-JEPA 2 is a “a powerful video-based world model that helps machines anticipate how the physical world behaves.” According to Meta, the first monetized application of this technology is probably going to be in Instagram Reels, where AI-powered scene descriptions may improve the accuracy of ad targeting.

Meta is also preparing for the surge in power demand that comes with AI. In a recent move that underscores its long-term energy planning, the company entered into a power agreement with Constellation Energy (CEG). This deal offers a significant hedge against energy price volatility while providing greater certainty in cost forecasting and operational planning.

Meta’s spending spree on acquisitions and innovation has not prevented it from delivering strong financial results. The company has raised its revenue and earnings at impressive compound annual growth rates (CAGRs) of 29% and 37%, respectively, over the past 10 years.

Moreover, the company has been on a streak of reporting earnings beats for the past nine quarters, with the most recent quarter seeing Meta reporting a beat on both revenue and earnings. Q1 2025 saw the company report revenues of $42.3 billion, up 16% from the previous year.

Earnings witnessed an even sharper rise of 37% to $6.43, coming in much ahead of the consensus estimate of an EPS of $5.24. The rise in revenue and earnings were also accompanied by an improvement in operating margins to 41% from 38% in the year-ago period.

Coming to cash flow, net cash from operating activities was at $24.03 billion, higher than the prior year’s figure of $19.25 billion. The company reported total cash and equivalents of $30.1 billion.

Overall, Meta expects to report revenues to be in the range of $42.5 billion to 45.5 billion in Q2 2025. The midpoint of this range denotes a yearly rise of 12.6%, even though the company lowered its revenue outlook for the full year to be between $113 billion and $118 billion, down from $114 billion to $119 billion.

Overall, analysts remain bullish about Meta stock, giving it a consensus rating of “Strong Buy” with a mean target price of $724.98. This denotes upside potential of about 1% from current levels. However, the high target price of $935 indicates upside potential of about 30% from current levels.

Out of 54 analysts covering the stock, 45 have a “Strong Buy” rating, three have a “Moderate Buy” rating, five have a “Hold” rating, and one has a “Strong Sell” rating.

On the date of publication, Pathikrit Bose did not have (either directly or indirectly) positions in any of the securities mentioned in this article. All information and data in this article is solely for informational purposes. This article was originally published on Barchart.com