MAROKO133 Update ai: People Really, Really Despise AI — Even More Than ICE, Poll Finds Edi

📌 MAROKO133 Breaking ai: People Really, Really Despise AI — Even More Than ICE, Po

Anti-AI sentiment surged over the last year as the hype surrounding the tech showed no signs of slowing down. Industry’s obsession with the tech has driven up electricity bills, been used to justify mass layoffs, and even helped the US military determine where to drop bombs on Iran.

It’s also quickly become an insufferable and practically inescapable part of everyday life, minting plenty of critics who range from average Americans to top AI researchers.

The backlash is enormous. According to a new national survey conducted by NBC News, AI is viewed even more negatively than the US Immigration and Customs Enforcement, the militarized agency that has been embroiled in major controversy over its brutal deportation program, including the fatal shooting of unarmed civilians.

According to the poll, only 26 percent of 1,000 registered voter respondents said they viewed AI positively, while a far bigger proportion of 46 percent viewed it negatively. In total, AI’s net favorability rating stands at a dismal negative 20 points.

Only the Democratic Party and Iran scored more negatively than AI, while ICE and Donald Trump scored slightly less terribly, with -18 and -12 points, respectively.

The shocking results once again highlight major disillusionment surrounding AI, an indictment underlining a growing schism between the excitement felt by company leaders and the quickly waning enthusiasm of their employees, who are being told to use the tech often against their will — potentially making their own roles redundant.

The optics of the Department of Defense employing AI to select targets in their bombing of Iran is certainly not helping, although it remains unclear how much the subject played a role for the poll respondents. The survey was conducted between February 27 and March 3, a date range that includes the beginning of Trump’s war on Iran. (That could also help explain Iran’s abysmal favorability rating of -53 points.)

Anthropic’s high-profile fight with the Department of Defense over where to draw the lines of ethical AI use in warfare, which culminated in the company suing the Pentagon today, also kicked off late last month in the days leading up to the start of the war.

Despite the pushback, Silicon Valley leaders and the Trump administration believe that AI represents the future. Tech giants continue to pour hundreds of billions of dollars into vast AI data center buildouts, which themselves have proven unpopular.

Besides allegedly skyrocketing energy bills, residents living near these facilities have been dealing with unbearable noise from gas-powered turbines brought in to cool the powerful AI chips.

More on AI sentiment: The Rage at OpenAI Has Grown So Immense That There Are Entire Protests Against It

The post People Really, Really Despise AI — Even More Than ICE, Poll Finds appeared first on Futurism.

🔗 Sumber: futurism.com

📌 MAROKO133 Eksklusif ai: MIT Researchers Unveil “SEAL”: A New Step Towards Self-I

The concept of AI self-improvement has been a hot topic in recent research circles, with a flurry of papers emerging and prominent figures like OpenAI CEO Sam Altman weighing in on the future of self-evolving intelligent systems. Now, a new paper from MIT, titled “Self-Adapting Language Models,” introduces SEAL (Self-Adapting LLMs), a novel framework that allows large language models (LLMs) to update their own weights. This development is seen as another significant step towards the realization of truly self-evolving AI.

The research paper, published yesterday, has already ignited considerable discussion, including on Hacker News. SEAL proposes a method where an LLM can generate its own training data through “self-editing” and subsequently update its weights based on new inputs. Crucially, this self-editing process is learned via reinforcement learning, with the reward mechanism tied to the updated model’s downstream performance.

The timing of this paper is particularly notable given the recent surge in interest surrounding AI self-evolution. Earlier this month, several other research efforts garnered attention, including Sakana AI and the University of British Columbia’s “Darwin-Gödel Machine (DGM),” CMU’s “Self-Rewarding Training (SRT),” Shanghai Jiao Tong University’s “MM-UPT” framework for continuous self-improvement in multimodal large models, and the “UI-Genie” self-improvement framework from The Chinese University of Hong Kong in collaboration with vivo.

Adding to the buzz, OpenAI CEO Sam Altman recently shared his vision of a future with self-improving AI and robots in his blog post, “The Gentle Singularity.” He posited that while the initial millions of humanoid robots would need traditional manufacturing, they would then be able to “operate the entire supply chain to build more robots, which can in turn build more chip fabrication facilities, data centers, and so on.” This was quickly followed by a tweet from @VraserX, claiming an OpenAI insider revealed the company was already running recursively self-improving AI internally, a claim that sparked widespread debate about its veracity.

Regardless of the specifics of internal OpenAI developments, the MIT paper on SEAL provides concrete evidence of AI’s progression towards self-evolution.

Understanding SEAL: Self-Adapting Language Models

The core idea behind SEAL is to enable language models to improve themselves when encountering new data by generating their own synthetic data and optimizing their parameters through self-editing. The model’s training objective is to directly generate these self-edits (SEs) using data provided within the model’s context.

The generation of these self-edits is learned through reinforcement learning. The model is rewarded when the generated self-edits, once applied, lead to improved performance on the target task. Therefore, SEAL can be conceptualized as an algorithm with two nested loops: an outer reinforcement learning (RL) loop that optimizes the generation of self-edits, and an inner update loop that uses the generated self-edits to update the model via gradient descent.

This method can be viewed as an instance of meta-learning, where the focus is on how to generate effective self-edits in a meta-learning fashion.

A General Framework

SEAL operates on a single task instance (C,τ), where C is context information relevant to the task, and τ defines the downstream evaluation for assessing the model’s adaptation. For example, in a knowledge integration task, C might be a passage to be integrated into the model’s internal knowledge, and τ a set of questions about that passage.

Given C, the model generates a self-edit SE, which then updates its parameters through supervised fine-tuning: θ′←SFT(θ,SE). Reinforcement learning is used to optimize this self-edit generation: the model performs an action (generates SE), receives a reward r based on LMθ′’s performance on τ, and updates its policy to maximize the expected reward.

The researchers found that traditional online policy methods like GRPO and PPO led to unstable training. They ultimately opted for ReST^EM, a simpler, filtering-based behavioral cloning approach from a DeepMind paper. This method can be viewed as an Expectation-Maximization (EM) process, where the E-step samples candidate outputs from the current model policy, and the M-step reinforces only those samples that yield a positive reward through supervised fine-tuning.

The paper also notes that while the current implementation uses a single model to generate and learn from self-edits, these roles could be separated in a “teacher-student” setup.

Instantiating SEAL in Specific Domains

The MIT team instantiated SEAL in two specific domains: knowledge integration and few-shot learning.

• Knowledge Integration: The goal here is to effectively integrate information from articles into the model’s weights.
• Few-Shot Learning: This involves the model adapting to new tasks with very few examples.

Experimental Results

The experimental results for both few-shot learning and knowledge integration demonstrate the effectiveness of the SEAL framework.

In few-shot learning, using a Llama-3.2-1B-Instruct model, SEAL significantly improved adaptation success rates, achieving 72.5% compared to 20% for models using basic self-edits without RL training, and 0% without adaptation. While still below “Oracle TTT” (an idealized baseline), this indicates substantial progress.

For knowledge integration, using a larger Qwen2.5-7B model to integrate new facts from SQuAD articles, SEAL consistently outperformed baseline methods. Training with synthetically generated data from the base Qwen-2.5-7B model already showed notable improvements, and subsequent reinforcement learning further boosted performance. The accuracy also showed rapid improvement over external RL iterations, often surpassing setups using GPT-4.1 generated data within just two iterations.

Qualitative examples from the paper illustrate how reinforcement learning leads to the generation of more detailed self-edits, resulting in improved performance.

While promising, the researchers also acknowledge some limitations of the SEAL framework, including aspects related to catastrophic forgetting, computational overhead, and context-dependent evaluation. These are discussed in detail in the original paper.

Original Paper: https://arxiv.org/pdf/2506.10943

Project Site: https://jyopari.github.io/posts/seal

Github Repo: https://github.com/Continual-Intelligence/SEAL

The post MIT Researchers Unveil “SEAL”: A New Step Towards Self-Improving AI first appeared on Synced.

🔗 Sumber: syncedreview.com