MAROKO133 Update ai: You’ll Be Sorry When You Hear What Justin Bieber’s $1.3 Million Bored

📌 MAROKO133 Eksklusif ai: You’ll Be Sorry When You Hear What Justin Bieber’s $1.3

In January 2022, when the world was still in the throes of the COVID-19 pandemic and coping with rolling lockdowns, non-fungible tokens were all the rage.

The blockchain-based assets, which more often than not took the form of cartoon pictures of silly-looking avatars like “CryptoPunks” to “Pudgy Penguins,” were selling like hot cakes. Even big shot celebrities were lining up to secure NFTs belonging to once-popular collections, like Yuga Labs’ Bored Ape Yacht Club.

In the midst of the craze, pop sensation Justin Bieber shelled out a hefty $1.3 million for a Bored Ape, an enormous sum of money for the rights to an image of a particularly glum-looking ape that appears to be on the verge of tears for some unknown reason.

Unsurprisingly, the questionable splurge turned out to be a hilariously bad investment. As Benzinga reports, the ape is now worth a measly $12,000, meaning that it’s lost over 99 percent of its value over the last three and change years.

The controversial crypto market has been going through an “NFT winter” following a brutal and extended crash. Collectors have gotten a hefty reality check. Who could’ve seen that coming?

Yuga Labs, the creator of the Bored Ape Yacht Club collection, has been holding on to dear life. The company has gone through several rounds of layoffs since the trend’s heyday, with Yuga Labs cofounder Greg Solano admitting in April 2024 that the company had “lost its way.”

In one particularly bizarre incident, partiers at a Bored Ape Yacht Club event in Hong Kong were alarmed after their eyes started burning, which later turned out to be caused by the event’s excessive use of UV light.

The company has also had to deal with a class action lawsuit that accused it of using celebrity endorsements to sell an unregistered security. The Securities Exchange Commission started an investigation into Yuga Labs in 2022, which concluded in March of last year, two months into Trump’s second term. The regulator’s takeaway was that NFTs weren’t securities after all. However, being let off the hook by regulators hasn’t exactly improved the situation, as more and more collectors are abandoning their NFTs.

Despite waning demand, Yuga Labs isn’t ready to call it quits. Earlier this year, the company announced it was looking to open an IRL Bored Ape clubhouse in Miami, featuring NFT galleries, event spaces, and “exclusive content” only members can explore.

But whether it can ever reignite the enormous amount of enthusiasm the trend once drew remains unclear at best.

While Bieber remains the proud owner of Bored Ape Yacht Club #3001, others are looking to cut their losses.

More on NFTs: Oops! The AWS Outage Took Down Everybody’s Bored Apes

The post You’ll Be Sorry When You Hear What Justin Bieber’s $1.3 Million Bored Ape Is Worth Now appeared first on Futurism.

🔗 Sumber: futurism.com

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

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