MAROKO133 Update ai: Details Emerge on Sam Altman’s Panic Sweats Hari Ini

📌 MAROKO133 Eksklusif ai: Details Emerge on Sam Altman’s Panic Sweats Wajib Baca

OpenAI has long made it its number one goal to realize artificial general intelligence, which it described in a 2023 blog post as “AI systems that are generally smarter than humans,” and which will benefit “all of humanity.”

Since then, experts have often accused the company of repeatedly shifting the goalposts, greatly watering down its original goal of an AI truly capable of surpassing the intellect of a human being.

And now, OpenAI CEO Sam Altman is reportedly setting aside what was once his firm’s top priority in an effort to stop the company from succumbing to its steep competition.

Last week, news emerged that the rattled executive had declared a “code red” in a note to staffers obtained by the Wall Street Journal, urging them to improve the quality of ChatGPT at the cost of delaying other projects, like advertising and a personal assistant.

Now, the newspaper has revealed new details about Altman’s call to arms, suggesting OpenAI “may have to pause” its quest to pursue AGI for the company to survive.

It’s a damning admission, highlighting how much pressure is building up on the company as it plans to spend well north of a trillion dollars to build out infrastructure over the next five years. Google, whose AI offerings are rapidly catching up, has clearly sent a strong signal, causing OpenAI’s executive branch to batten up the hatches and double down on its core offering, ChatGPT.

Instead of vetting the tool’s output with the help of human professionals, Altman is looking to make “better use of user signals,” per the WSJ. In other words, the company is doubling down on user feedback to boost engagement — even if that means making its models more sycophantic, which can have disastrous side effects.

It’s a neck-in-neck race between OpenAI and Google. OpenAI is expected to release its latest AI model, called 5.2, later this week, likely a response to Google’s Gemini 3, which impressed with benchmarks that exceeded OpenAI’s current most powerful models.

Google’s Nano Banana Pro AI image model, which was released last month, has also been hailed as a substantial leap, while OpenAI’s video and controversy-generating app, Sora, has fallen by the wayside. In fact, according to the WSJ, Sora may also be put on pause as OpenAI doubles down on ChatGPT.

OpenAI staffers appear to be painfully aware of OpenAI and Google trading blows, closely following LM Arena, an AI leaderboard that assigns each AI model a score based on users choosing the best output to the same prompt between two AI models.

Indeed, Altman argued in his memo that “we should be at the top of things like LM [A]rena.”

To do so, the executive is calling on the company to focus on making its AI models more personable, a quality that experts warn could lead to more users spiraling into severe delusions.

Where that leaves OpenAI’s original goal of building an AI that can surpass the intellect of a human being remains unclear at best. Altman, who has long garnered a reputation for setting sweeping and extremely ambitious goals, is now singing a notably different tune from before — as his company doubles down on its number one money maker at all costs.

More on Altman: OpenAI Is Suddenly in Major Trouble

The post Details Emerge on Sam Altman’s Panic Sweats appeared first on Futurism.

🔗 Sumber: futurism.com

📌 MAROKO133 Update ai: Researchers from PSU and Duke introduce “Multi-Agent System

Share My Research is Synced’s column that welcomes scholars to share their own research breakthroughs with over 2M global AI enthusiasts. Beyond technological advances, Share My Research also calls for interesting stories behind the research and exciting research ideas. 

Meet the author
Institutions: Penn State University, Duke University, Google DeepMind, University of Washington, Meta, Nanyang Technological University, and Oregon State University. The co-first authors are Shaokun Zhang of Penn State University and Ming Yin of Duke University.

In recent years, LLM Multi-Agent systems have garnered widespread attention for their collaborative approach to solving complex problems. However, it’s a common scenario for these systems to fail at a task despite a flurry of activity. This leaves developers with a critical question: which agent, at what point, was responsible for the failure? Sifting through vast interaction logs to pinpoint the root cause feels like finding a needle in a haystack—a time-consuming and labor-intensive effort.
 
This is a familiar frustration for developers. In increasingly complex Multi-Agent systems, failures are not only common but also incredibly difficult to diagnose due to the autonomous nature of agent collaboration and long information chains. Without a way to quickly identify the source of a failure, system iteration and optimization grind to a halt.
 
To address this challenge, researchers from Penn State University and Duke University, in collaboration with institutions including Google DeepMind, have introduced the novel research problem of “Automated Failure Attribution.” They have constructed the first benchmark dataset for this task, Who&When, and have developed and evaluated several automated attribution methods. This work not only highlights the complexity of the task but also paves a new path toward enhancing the reliability of LLM Multi-Agent systems.
The paper has been accepted as a Spotlight presentation at the top-tier machine learning conference, ICML 2025, and the code and dataset are now fully open-source.

Paper：https://arxiv.org/pdf/2505.00212
Code：https://github.com/mingyin1/Agents_Failure_Attribution
Dataset：https://huggingface.co/datasets/Kevin355/Who_and_When
 
 
Research Background and Challenges
LLM-driven Multi-Agent systems have demonstrated immense potential across many domains. However, these systems are fragile; errors by a single agent, misunderstandings between agents, or mistakes in information transmission can lead to the failure of the entire task.

Currently, when a system fails, developers are often left with manual and inefficient methods for debugging:
Manual Log Archaeology : Developers must manually review lengthy interaction logs to find the source of the problem.
Reliance on Expertise : The debugging process is highly dependent on the developer’s deep understanding of the system and the task at hand.
 
This “needle in a haystack” approach to debugging is not only inefficient but also severely hinders rapid system iteration and the improvement of system reliability. There is an urgent need for an automated, systematic method to pinpoint the cause of failures, effectively bridging the gap between “evaluation results” and “system improvement.”

Core Contributions
This paper makes several groundbreaking contributions to address the challenges above:
1. Defining a New Problem: The paper is the first to formalize “automated failure attribution” as a specific research task. This task is defined by identifying the

2. failure-responsible agent and the decisive error step that led to the task’s failure.

Constructing the First Benchmark Dataset: Who&When : This dataset includes a wide range of failure logs collected from 127 LLM Multi-Agent systems, which were either algorithmically generated or hand-crafted by experts to ensure realism and diversity. Each failure log is accompanied by fine-grained human annotations for:
Who: The agent responsible for the failure.
When: The specific interaction step where the decisive error occurred.
Why: A natural language explanation of the cause of the failure.

3. Exploring Initial “Automated Attribution” Methods : Using the Who&When dataset, the paper designs and assesses three distinct methods for automated failure attribution:
All-at-Once: This method provides the LLM with the user query and the complete failure log, asking it to identify the responsible agent and the decisive error step in a single pass. While cost-effective, it may struggle to pinpoint precise errors in long contexts.
Step-by-Step: This approach mimics manual debugging by having the LLM review the interaction log sequentially, making a judgment at each step until the error is found. It is more precise at locating the error step but incurs higher costs and risks accumulating errors.
Binary Search: A compromise between the first two methods, this strategy repeatedly divides the log in half, using the LLM to determine which segment contains the error. It then recursively searches the identified segment, offering a balance of cost and performance.
 
Experimental Results and Key Findings

Experiments were conducted in two settings: one where the LLM knows the ground truth answer to the problem the Multi-Agent system is trying to solve (With Ground Truth) and one where it does not (Without Ground Truth). The primary model used was GPT-4o, though other models were also tested. The systematic evaluation of these methods on the Who&When dataset yielded several important insights:

• A Long Way to Go: Current methods are far from perfect. Even the best-performing single method achieved an accuracy of only about 53.5% in identifying the responsible agent and a mere 14.2% in pinpointing the exact error step. Some methods performed even worse than random guessing, underscoring the difficulty of the task.
• No “All-in-One” Solution: Different methods excel at different aspects of the problem. The All-at-Once method is better at identifying “Who,” while the Step-by-Step method is more effective at determining “When.” The Binary Search method provides a middle-ground performance.

• Hybrid Approaches Show Promise but at a High Cost: The researchers found that combining different methods, such as using the All-at-Once approach to identify a potential agent and then applying the Step-by-Step method to find the error, can improve overall performance. However, this comes with a significant increase in computational cost.

• State-of-the-Art Models Struggle: Surprisingly, even the most advanced reasoning m…

  Konten dipersingkat otomatis.

  🔗 Sumber: syncedreview.com

  ---