Connect with us

Brand Stories

OpenAI’s artificial intelligence (AI) model has achieved a monumental achievement in catching up wit..

Published

on


Using a closed inference model under experimentation in my department, I scored 35 out of 42 points in IMO 2025 and have struggled with AI for a long time given the same time as humans…10 to 30% accuracy for other models

The closed inference giant language model (LLM), which OpenAI is experimenting internally, has achieved the equivalent of a gold medal at the International Mathematical Olympiad (IMO). [Source = Alexander OpenAI Research Scientist X]

OpenAI’s artificial intelligence (AI) model has achieved a monumental achievement in catching up with human intelligence at the International Mathematical Olympiad.

In other words, OpenAI has shown its robustness with overwhelming AI performance at a time when rumors of a crisis are emerging due to the failure to acquire start-up Windsurf and the subsequent outflow of talent.

“We achieved gold medal-level performance at the International Mathematical Olympiad (IMO 2025) this year with a universal inference model,” OpenAI CEO Sam Altman said on his X(X) on the 19th (local time).

“When I first started OpenAI, it was a dream story. This is an important indicator of how advanced AI has developed over the past decade, he said, explaining the significance of this achievement.

The results are based on the Large Language Model for Inference (LLM), which is being experimented internally by a small team led by Alexander Way, a research scientist at OpenAI.

IMO is a prestigious Olympiad that has been underway since 1959, and students under the age of 20 representing each country participate. It is characterized by requiring mathematical thinking and creative ideas, not just problems that can be solved by memorizing formulas.

According to OpenAI, the test was conducted under the same conditions as human candidates. IMO, which consists of a total of 6 questions, is a method of solving 3 questions for 4 hours and 30 minutes a day over 2 days.

OpenAI’s model scored 35 points out of 42 while solving 5 out of 6 questions, setting a record equivalent to the gold medal spot.

In this year’s IMO, humans still performed better than AI with six perfect scores, but it is evaluated as a symbolic event that shows how close LLM’s rapidly developing performance is to human intelligence.

LLMs so far have hardly reached the silver and bronze medals in IMO, let alone the gold medal.

Google DeepMind’s “AlphaProof” and “AlphaGeometry 2” scored silver medals last year, but these models were specialized only in the math field.

Noam Brown, an open AI research scientist, said, “The achievements that AI has previously shown in Go and poker are the result of researchers training AI to master only that specific area for years. However, this model is not an IMO-specific model, but an inference LLM that combines a new method of experimental general-purpose technology.”

According to MathArena of the Federal Institute of Technology (ETH Zurich), Switzerland, which tracks the mathematical performance of major models, all of the world’s leading models such as Google’s Gemini 2.5 Pro, xAI’s Grok 4 and DeepSeek’s R1 did not even make it to the bronze medal list at this year’s IMO 2025.

The specific secret of OpenAI’s breakthrough achievement has not been disclosed.

“We have developed a new technology that enables LLM to perform much better tasks that are difficult to verify,” Brown said. “O1 (existing inference model) thinks for seconds, and ‘deep research’ functions take minutes, but this model thinks for hours.”

Meanwhile, this result is not a third-party verification state as it is conducted as a closed-door experimental model that has not been officially released by OpenAI. Mass Arena said in response, “We are excited to see steep progress in this area and look forward to the launch of the model, enabling independent evaluation through public benchmarks.”



Source link

Continue Reading
Click to comment

You must be logged in to post a comment Login

Leave a Reply

Brand Stories

Artificial intelligence: Israel backs NIS 1m. fund for sandboxes – The Jerusalem Post

Published

on

By



Artificial intelligence: Israel backs NIS 1m. fund for sandboxes  The Jerusalem Post



Source link

Continue Reading

Brand Stories

5 Artificial Intelligence (AI) Infrastructure Stocks Powering the Next Wave of Innovation

Published

on


Demand for AI computing power could push data center spending to nearly $7 trillion by 2030.

It will be a massive undertaking to build out the hardware and support necessary to power increasingly advanced artificial intelligence and provide it at a global level where billions of people can access it.

According to research by McKinsey & Company, the world’s technology needs will require $6.7 trillion in data center spending by 2030. Of that, $5 trillion will be due to the rising processing power demands of artificial intelligence (AI). These investments, though, will lay the groundwork for the next era of global innovation, which will revolutionize existing industries and create new ones.

Some key companies have already been experiencing significant growth due to the AI trend, and there is still likely a long runway ahead for players in key AI infrastructure spaces, including semiconductors, cloud computing, and networking.

Here are five top stocks to buy and hold for the next wave of AI innovation.

Image source: GETTY IMAGES

Nvidia: The data center AI chip leader

Inside these colossal AI data centers are many thousands of AI accelerator chips, usually from Nvidia (NVDA -0.42%). The company’s graphics processing units (GPUs) are the only ones that can make use of its proprietary CUDA platform, which contains an array of tools and libraries to help developers build and deploy applications that use the hardware efficiently. CUDA’s effectiveness — and its popularity with developers — has helped Nvidia win an estimated 92% share of the data center GPU market.

The company has maintained its winning position as it progressed from its previous Hopper architecture to its current Blackwell chips, and it expects to launch its next-generation architecture, with a CPU called Vera and a GPU called Rubin, next year. Analysts expect Nvidia’s revenue to grow to $200 billion this year and $251 billion in 2026.

Amazon and Microsoft: Winning in AI through the cloud

AI software is primarily trained and powered through large cloud data centers, making the leading cloud infrastructure companies vital pieces of the equation. They’re also Nvidia’s largest customers. Amazon (AMZN 0.97%) Web Services (AWS) has long been the world’s leading cloud platform, with about 30% of the cloud infrastructure market today.Through the cloud, companies can access and deploy AI agents, models, and other software throughout their businesses. 

AWS’s sales grew by 17% year over year in Q1, and it should maintain a similar pace. Goldman Sachs estimates that AI demand will drive cloud computing sales industrywide to $2 trillion by 2030. Amazon will capture a significant portion of that, and since AWS is Amazon’s primary profit center, the company’s bottom line should also thrive.

It’s a similar theme for Microsoft (MSFT -0.33%). Its Azure is the world’s second-largest cloud platform, with a market share of approximately 21%. Microsoft stands out from the pack for its deep ties with millions of corporate clients. Businesses rely on Microsoft’s range of hardware and software products, including its enterprise software, the Windows operating system, and productivity applications such as Outlook and Excel.

Microsoft’s vast ecosystem creates sticky revenue streams and provides it with an enormous customer base to cross-sell its AI products and services to. Microsoft has also invested in OpenAI, the developer behind ChatGPT, and works with it extensively, although that relationship has become somewhat strained as OpenAI has grown increasingly successful.

Regardless, Microsoft’s massive footprint across the AI and broader tech space makes it a no-brainer.

Arista Networks and Broadcom: The networking tech that underpins AI

Within data centers, huge clusters of AI chips must communicate and work together, which requires them to transfer massive amounts of data at extremely high speeds. Arista Networks (ANET -0.06%) sells high-end networking switches and software that help accomplish this. The company has already thrived in this golden age of data centers, with top clients including Microsoft and Meta Platforms, which happen to also be among the highest spenders on AI infrastructure.

Arista Networks will likely continue benefiting from growth in AI investments, as these increasingly powerful AI models consume ever-increasing amounts of data. Analysts expect Arista Networks to generate $8.4 billion in sales this year (versus $7 billion last year), then $9.9 billion next year, with nearly 19% annualized long-term earnings growth.

Tightly woven into this same theme is Broadcom (AVGO -1.12%), which specializes in designing semiconductors used for networking applications.

For example, Arista Networks utilizes Broadcom’s Tomahawk and Jericho silicon in the networking switches it builds for data centers. Broadcom’s AI-related semiconductor sales increased by 46% year-over-year in the second quarter.

Looking further out, Broadcom is becoming a more prominent role player in AI infrastructure. It has designed custom accelerator chips (XPUs) for AI model training and inference. It has struck partnerships with at least three AI customers that management believes will each deploy clusters of 1 million accelerator chips by 2027. Broadcom’s red-hot AI momentum has analysts estimating the company will grow earnings by an average of 23% annually over the next three to five years.

Randi Zuckerberg, a former director of market development and spokeswoman for Facebook and sister to Meta Platforms CEO Mark Zuckerberg, is a member of The Motley Fool’s board of directors. John Mackey, former CEO of Whole Foods Market, an Amazon subsidiary, is a member of The Motley Fool’s board of directors. Justin Pope has no position in any of the stocks mentioned. The Motley Fool has positions in and recommends Amazon, Arista Networks, Goldman Sachs Group, Meta Platforms, Microsoft, and Nvidia. The Motley Fool recommends Broadcom and recommends the following options: long January 2026 $395 calls on Microsoft and short January 2026 $405 calls on Microsoft. The Motley Fool has a disclosure policy.



Source link

Continue Reading

Brand Stories

MAGA AI bot network divided on Trump-Epstein backlash

Published

on


A previously unreported network of hundreds of accounts on X is using artificial intelligence to automatically reply to conservatives with positive messages about people in the Trump administration, researchers say.

But with the MAGA movement split over the administration’s handling of files involving deceased sex offender Jeffrey Epstein, the accounts’ messaging has broken, offering contradictory statements on the issue and revealing the LLM-fueled nature of the accounts.

The network, tracked for NBC News by both the social media analytics company Alethea and researchers at Clemson University, consists of more than 400 identified bot accounts, though the number could be far larger, the researchers say. Its accounts offer consistent praise for key Trump figures, particularly support for Health Secretary Robert F. Kennedy Jr. and White House press secretary Karoline Leavitt.

As often is the case with bot accounts, those viewed by NBC News tended to have only a few dozen followers, and their posts rarely get many views. But a large audience does not appear to be the point. Their effectiveness, if they have any, is in the hope that they contribute to a partisan echo chamber, and that en masse they can “massage perceptions,” said Darren Linvill, the director of Clemson University’s Media Forensics Hub, which studies online disinformation campaigns.

“They’re not really there to get engagement. They’re there to just be occasionally seen in those replies,” Linvill told NBC News.

The researchers declined to share specifics on how they identified the accounts, but noted they shared a number of distinct trends. All were created, seemingly in batches, around three specific days last year. They frequently punctuate their posts with hashtags, often ones that are irrelevant to the conversation. They post almost exclusively by replying to other users, often to people who pay X for verification and by repeating similarly worded sentiments over and over in short succession. At times, they will respond to someone’s post by repeating it back to them verbatim.

It’s unclear who is behind the network, or which of the multiple AI chatbots that are widely accessible to the public was used to power it.

The bots have posted support for conservative figures since 2024, including supporting Trump and other Republicans on the ballot in the lead-up to the election, and then afterward posting that they were excited for Trump to take office. Though they would occasionally mix their messages — some have professed affection for MSNBC host Rachel Maddow, for instance — their messaging was consistently in favor of MAGA figures until the recent Epstein files controversy.

A core constituency of Trump supporters voted for him on the belief that Trump, a former friend of Epstein’s, would expose a list of supposed rich and powerful clients and bring justice to Epstein’s victims.

It’s only since earlier this month, when Attorney General Pam Bondi announced she would not release additional Epstein files, that the accounts’ messaging has become so split, with some accounts telling different users opposite opinions almost concurrently.

During the same minute last Saturday morning, for example, one account in the network both cautioned a MAGA supporter from judging Bondi too harshly and told another that Bondi or FBI Director Kash Patel and Deputy Director Dan Bongino should resign over the scandal.



Source link

Continue Reading

Trending

Copyright © 2025 travelstoriz.com.