{"pageProps":{"settings":{"title":"The Batch | DeepLearning.AI","description":"Weekly AI news for engineers, executives, and enthusiasts.","logo":"https://charonhub.deeplearning.ai/content/images/2021/04/LogoFiles_DeepLearning_PrimaryLogo.png","icon":"https://charonhub.deeplearning.ai/content/images/size/w256h256/2021/05/logo.png","accent_color":"#F65B66","cover_image":"https://charonhub.deeplearning.ai/content/images/2021/04/v2osk-Ovn1hyBge38-unsplash.jpg","facebook":"DeepLearningAIHQ/","twitter":"@DeepLearningAI","locale":"en","timezone":"America/Los_Angeles","codeinjection_head":null,"codeinjection_foot":null,"navigation":[],"secondary_navigation":[],"meta_title":"The Batch | DeepLearning.AI","meta_description":"Weekly AI news for engineers, executives, and enthusiasts.","og_image":"https://charonhub.deeplearning.ai/content/images/2021/08/thebatchlogotwitter-1.png","og_title":"The Batch | DeepLearning.AI","og_description":"Weekly AI news for engineers, executives, and enthusiasts.","twitter_image":"https://charonhub.deeplearning.ai/content/images/2021/08/thebatchlogotwitter.png","twitter_title":"The Batch | DeepLearning.AI","twitter_description":"Weekly AI news for engineers, executives, and enthusiasts.","members_support_address":"noreply","members_enabled":true,"allow_self_signup":false,"members_invite_only":true,"members_signup_access":"invite","paid_members_enabled":false,"firstpromoter_account":null,"portal_button_style":"icon-and-text","portal_button_signup_text":"Subscribe","portal_button_icon":"icon-3","portal_signup_terms_html":null,"portal_signup_checkbox_required":false,"portal_plans":["free","monthly","yearly"],"portal_default_plan":"yearly","portal_name":false,"portal_button":true,"comments_enabled":"off","recommendations_enabled":false,"outbound_link_tagging":false,"default_email_address":"noreply@charonhub.deeplearning.ai","support_email_address":"noreply@charonhub.deeplearning.ai","editor_default_email_recipients":"disabled","labs":{},"lang":"en","url":"https://charonhub.deeplearning.ai","version":"5.120"},"post":{"slug":"issue-326","id":"690baffbb66ce900018a0df8","uuid":"7c90b428-7576-45fe-9a92-ead2e84061e7","title":"OpenAI Reorgs For Profit, MiniMax-M2 Leads Open Coding, Universal Music Group Embraces AI, LLMs Go Private","html":"\n\n

\n\n

Dear friends,

AI agents are getting better at looking at different types of data in businesses to spot patterns and create value. This is making data silos increasingly painful. This is why I increasingly try to select software that lets me control my own data, so I can make it available to my AI agents.

Because of AI’s growing capabilities, the value you can now create from “connecting the dots” between different pieces of data is higher than ever. For example, if an email click is logged in one vendor’s system and a subsequent online purchase is logged in a different one, then it is valuable to build agents that can access both of these data sources to see how they correlate to make better decisions.

Unfortunately, many SaaS vendors try to create a data silo in their customer’s business. By making it hard for you to extract your data, they create high switching costs. This also allows them to steer you to buy their AI agent services — sometimes at high expense and/or of low quality — rather than build your own or buy from a different vendor. Unfortunately, some SaaS vendors are seeing AI agents coming for this data and working to make it harder for you (and your AI agents) to efficiently access it.

One of my teams just told me that a SaaS vendor we have been using to store our customer data wants to charge over $20,000 for an API key to get at our data. This high cost — no doubt intentionally designed to make it hard for customers to get their data out — is adding a barrier to implementing agentic workflows that take advantage of that data.

Through AI Aspire (an AI advisory firm), I occasionally advise businesses on their AI strategies. When it comes to buying SaaS, I often advise them to try to control their own data (which, sadly, some vendors mightily resist). This way, you can hire a SaaS vendor to record and operate on your data, but ultimately you decide how to route it to the appropriate human or AI system for processing.

Over the past decade, a lot of work has gone into organizing businesses’ structured data. Because AI can now process unstructured data much better than before, the value of organizing your unstructured data (including PDF files, which LandingAI’s Agentic Document Extraction specializes in!) is higher than ever before.

In the era of generative AI, businesses and individuals have important work ahead to organize their data to be AI-ready.

Keep building,

Andrew

P.S. As an individual, my favorite note-taking app is Obsidian. I am happy to “hire” Obsidian to operate on my notes files. And, all my notes are saved as Markdown files in my file system, and I have built AI agents that read from or write to my Obsidian files. This is a small example of how controlling my own notes data lets me do more with AI agents!

A MESSAGE FROM DEEPLEARNING.AI

In Jupyter AI: AI Coding in Notebooks, Andrew Ng and Brian Granger (co-founder of Project Jupyter) teach you to use Jupyter AI as your notebook coding partner. Learn to generate, refactor, and debug code through an integrated chat interface. Available now at DeepLearning.AI. Enroll today!

News

OpenAI Reorganizes For Profit

OpenAI completed its transition from nonprofit to for-profit in a feat of legal engineering that took an army of lawyers, investment bankers, and two state attorneys general 18 months to negotiate.

What’s new: The restructured OpenAI Group PBC is a public benefit corporation, a for-profit company with a mission to create a positive social impact. It can earn unlimited returns to its investors, which clears the way to attract further investments including a possible initial public offering. It remains overseen by a nonprofit foundation, the newly renamed Open AI Foundation, which owns a 26 percent stake in the corporation. Microsoft holds a 27 percent stake in OpenAI under new terms for the companies’ partnership.

How it works: The agreement frees OpenAI from the constraints of its 2015 nonprofit beginnings that have limited investors to a 100x maximum return since an earlier restructuring in 2019. The new structure aims to satisfy the concerns of state officials in California and Delaware that the old structure created conflicts of interest between serving the public and rewarding shareholders, even as it aims to preserve the company’s mission to ensure that artificial general intelligence (AGI), if and when OpenAI builds it, will benefit humanity.

holds

Bloomberg

Behind the news: This isn’t the restructuring OpenAI originally wanted. A 2024 plan would have eliminated the nonprofit and turned the company into a traditional venture-backed entity. The California and Delaware attorneys general balked at that proposal, which led to a compromise that keeps the nonprofit in charge.

OpenAI needed California’s and Delaware’s approvals to avoid losing a $40 billion investment from SoftBank, half of which was contingent on the restructuring and lifting of the cap on investor returns. It also needed Microsoft’s agreement. This gave Microsoft significant leverage over the terms.OpenAI committed to remaining in California, and thus to continue to be subject to the state’s oversight, as part of its negotiation with California’s attorney general.

Why it matters: OpenAI has achieved staggering growth in its user base and valuation in spite of its nonprofit status. The new restructure adds pressure to get on a road to profitability. The company’s annual revenue run rate reportedly is greater than $13 billion, but given its commitment to spend an estimated $1 trillion on computing infrastructure, further funding is necessary to finance its ambitions.

We’re thinking: Microsoft’s early investments in OpenAI have more than paid off. When Microsoft CEO Satya Nadella proposed his company’s initial $1 billion investment in 2019, Bill Gates warned, “You’re going to burn this billion dollars.” Microsoft’s total investment of $13 billion is now worth $135 billion.

Open-Weights Coding Leader

An open-weights model from Shanghai-based MiniMax challenges top proprietary models on key benchmarks for coding and agentic tasks.

What’s new: MiniMax, which provides voice-chat and image-generation services, released the weights for MiniMax-M2, a large language model that’s optimized for coding and agentic tasks.

Input/output:

204,000

Architecture:

Performance:

Availability:

Hugging Face

ModelScope

MiniMax

Undisclosed:

How it works: MiniMax has not published a technical report on MiniMax-M2, so little public information is available about how it built the model.

Given a prompt, MiniMax-M2 interleaves reasoning steps (enclosed within <think>...</think> tags) within its output. This differs from models like DeepSeek-R1 that generate a block of reasoning steps prior to final output. It also differs from models like OpenAI GPT-5 and recent Anthropic Claude models that also generate reasoning steps prior to final output but hide or summarize them.MiniMax advises users to retain <think>...</think> tags in their conversation histories for optimal performance across multiple turns, because removing them (say, to economize on tokens) would degrade the model’s context.

Results: MiniMax-M2 achieved 61 on independent evaluator Artificial Analysis’ Intelligence Index (a weighted average of benchmark performance in mathematics, science, reasoning, and coding), a new high for open weights models, ahead of DeepSeek-V3.2 (57 points) and Kimi K2 (50 points). It trails proprietary models GPT-5 with thinking enabled (69 points) and Claude Sonnet 4.5 (63 points). Beyond that, it excelled in coding and agentic tasks but proved notably verbose. It consumed 120 million tokens to complete Artificial Analysis evaluations, tied for highest with Grok 4.

Behind the news: In June, MiniMax published weights for MiniMax-M1, a reasoning model designed to support agentic workflows over long contexts (1 million tokens). The company had been developing agents for internal use in tasks like coding, processing user feedback, and screening resumes. However, it found that leading closed-weights models were too costly and slow, while open-weights alternatives were less capable. It says it built MiniMax-M2 to fill the gap.

Why it matters: Developing reliable agentic applications requires experimenting with combinations and permutations of prompts, tools, and task decompositions, which generates lots of tokens. Cost-effective models that are capable of agentic tasks, like MiniMax-M2, can help more small teams innovate with agents.

We’re thinking: MiniMax-M2’s visible reasoning traces make its decisions more auditable than models that hide or summarize their reasoning steps. As agents are applied increasingly to mission-critical applications, transparency in reasoning may matter as much as raw performance.

AI Music With Major-Label Support

Music-generation service Udio will build an AI streaming platform in collaboration with the world’s biggest record label.

What’s new: Udio plans to launch a paid platform that enables fans to generate music based on recordings by artists on Universal Music Group (BMG) and its subsidiary labels. UMG artists include Taylor Swift, Olivia Rodrigo, Kendrick Lamar, and many other best-selling performers. The venture is part of an agreement to settle a lawsuit filed last year, in which UMG alleged that Udio had violated its copyrights when the AI company trained its music models. The financial terms, duration, and the remainder of the settlement are undisclosed. Udio is free to make similar arrangements with other music labels, the music-industry news publication Billboard reported.

How it works: The platform will allow paying customers to remix, customize, and combine existing recordings and share them with other subscribers.

restored

Other deals: In addition to Udio, UMG forged relationships with other AI music companies that supply tools and technology.

said

announced

Behind the news: Like book publishers and movie studios, recording companies have moved aggressively to stop AI companies from training models on materials they control and generating output that might compete with them.

devised

sued

counter

Why it matters: Music labels, like other media companies, see their businesses threatened by generative AI, which can synthesize products that are superficially similar to their own at lower cost and in less time. A study by the French streaming music service Deezer found that nearly 28 percent of the music it delivered was generated. In June, a musical group called Velvet Sundown racked up 1 million plays on Spotify of music generated by Suno. The settlement between Udio and UMG unites traditional and AI-generated music in a single business and suggests there could be common ground between media and AI companies, albeit with side effects such as limiting Udio’s distribution of generated music.

We’re thinking: Lawsuits against Suno and Udio by Sony Music, Warner Music, and the RIAA are still underway. This deal offers a blueprint for resolving those cases, but their outcomes are by no means certain. As lovers of music, we look forward to hearing more of it.

Masking Private Data in Training Sets

Large language models often memorize details in their training data, including private information that may appear only once, like a person’s name, address, or phone number. Researchers built the first open-weights language model that’s guaranteed not to remember such facts.

What’s new: Amer Sinha, Thomas Mesnard, and colleagues at Google released VaultGemma, a 1 billion-parameter model that’s trained from scratch using the technique known as differential privacy. This method prevented the model from memorizing examples that occurred only once in its training data, with a modest sacrifice in performance. Weights are free to download under a license that allows noncommercial and commercial uses with some restrictions.

Differential privacy basics: An algorithm (such as training a neural network) is differentially private if it’s impossible to tell the difference between its product (the learned weights) and its product given that same dataset minus any given example. Since the presence or absence of a single example can’t significantly change the product, personal information can’t leak from the product (the model’s weights) or the consequences of the product (the model’s outputs). In training a neural network, it’s possible to limit the impact of one example by limiting how much each example’s gradient can impact a model’s weights, for instance, by adding noise to each example’s gradient to make it harder to tell from that of other examples.

Key insight: Most previous work applies differential privacy when fine-tuning a model, but that doesn’t prevent the model from memorizing an example during pretraining. Once private information is encoded in the model’s weights, later fine-tuning can’t remove it reliably. Training with differential privacy from the start ensures that such details don’t become embedded in the model.

How it works: VaultGemma follows the same transformer architecture as the 1 billion-parameter version of Google’s Gemma 2. Moreover, the authors pretrained it from scratch on the same 13-trillion-token dataset as Gemma 2 (web, code, and scientific text). The authors applied differential privacy as VaultGemma learned to predict the next token in sequences of 1,024 tokens.

For every example in a batch, the authors computed the gradient and clipped it so its contribution to the weight update didn’t exceed a fixed threshold. This ensured that any given example didn’t have a disproportionate impact on the weight updates relative to other examples.The authors averaged the clipped gradients across each batch and added Gaussian noise to the average before updating the model’s weights. The noise weakened the influence of unique examples while allowing repeated examples to stand out. As a result, the model’s weights became statistically indistinguishable from those of a model trained without any particular 1,024-token sequence.

Results: VaultGemma showed no measurable memorization across 1 million sequences sampled at random from the training set, while pretrained Gemma 1, 2, and 3 models of roughly similar sizes did. VaultGemma’s average performance across 7 question-answering benchmarks generally matched that of GPT-2 (1.5 billion parameters) but fell short of Gemma models of roughly similar size.

HellaSwag

TriviaQA

Yes, but: The privacy protection comes with a caveat: It applies only to unique examples such as a private phone number that occurs only once in the dataset. If private information appears repeatedly in the training data, for instance, a celebrity's street address that leaked and appeared in several publications, the model can learn it as a general pattern.

Why it matters: Private information can find its way into training datasets, and in normal use, a typical large language model may divulge it without the subject’s consent. VaultGemma shows that large open-weights models can be provably private. While such privacy still comes with a cost — VaultGemma 1B performs roughly on par with models built about five years ago — the results are promising, and future work may close that gap.

We’re thinking: The smartest model for sensitive data might be the one that remembers only the most common information.

","comment_id":"690baffbb66ce900018a0df8","feature_image":"https://charonhub.deeplearning.ai/content/images/2025/11/Tear-Down-Data-Silos--1.png","featured":false,"visibility":"public","created_at":"2025-11-05T12:13:47.000-08:00","updated_at":"2025-11-05T12:34:21.000-08:00","published_at":"2025-11-05T12:32:00.000-08:00","custom_excerpt":"The Batch AI News and Insights: AI agents are getting better at looking at different types of data in businesses to spot patterns and create value. This is making data silos increasingly painful.","codeinjection_head":null,"codeinjection_foot":null,"custom_template":null,"canonical_url":null,"tags":[{"id":"684b8da3e105f4000191bb07","name":"The Batch Newsletter","slug":"the-batch","description":"Weekly AI news and perspective for engineers, executives, and enthusiasts.","feature_image":null,"visibility":"public","og_image":"https://charonhub.deeplearning.ai/content/images/2021/08/thebatchlogotwitter-3.png","og_title":"The Batch Newsletter | DeepLearning.AI","og_description":"Weekly AI news for engineers, executives, and enthusiasts.","twitter_image":"https://charonhub.deeplearning.ai/content/images/2021/08/thebatchlogotwitter-2.png","twitter_title":"The Batch Newsletter | DeepLearning.AI","twitter_description":"Weekly AI news for engineers, executives, and enthusiasts.","meta_title":"The Batch Newsletter | DeepLearning.AI","meta_description":"Weekly AI news for engineers, executives, and enthusiasts.","codeinjection_head":null,"codeinjection_foot":null,"canonical_url":null,"accent_color":null,"url":"https://charonhub.deeplearning.ai/tag/the-batch/"},{"id":"690bb4a55a53c50001015f69","name":"Nov 05, 2025","slug":"nov-05-2025","description":null,"feature_image":null,"visibility":"public","og_image":null,"og_title":null,"og_description":null,"twitter_image":null,"twitter_title":null,"twitter_description":null,"meta_title":null,"meta_description":null,"codeinjection_head":null,"codeinjection_foot":null,"canonical_url":null,"accent_color":null,"url":"https://charonhub.deeplearning.ai/tag/nov-05-2025/"},{"id":"690bb4a55a53c50001015f6a","name":"issue-326","slug":"issue-326","description":null,"feature_image":null,"visibility":"public","og_image":null,"og_title":null,"og_description":null,"twitter_image":null,"twitter_title":null,"twitter_description":null,"meta_title":null,"meta_description":null,"codeinjection_head":null,"codeinjection_foot":null,"canonical_url":null,"accent_color":null,"url":"https://charonhub.deeplearning.ai/tag/issue-326/"}],"authors":[{"id":"1","name":"Analytics DeepLearning.AI","slug":"analytics","profile_image":null,"cover_image":null,"bio":null,"website":null,"location":null,"facebook":null,"twitter":null,"threads":null,"bluesky":null,"mastodon":null,"tiktok":null,"youtube":null,"instagram":null,"linkedin":null,"meta_title":null,"meta_description":null,"url":"https://charonhub.deeplearning.ai/author/analytics/"}],"primary_author":{"id":"1","name":"Analytics DeepLearning.AI","slug":"analytics","profile_image":null,"cover_image":null,"bio":null,"website":null,"location":null,"facebook":null,"twitter":null,"threads":null,"bluesky":null,"mastodon":null,"tiktok":null,"youtube":null,"instagram":null,"linkedin":null,"meta_title":null,"meta_description":null,"url":"https://charonhub.deeplearning.ai/author/analytics/"},"primary_tag":{"id":"684b8da3e105f4000191bb07","name":"The Batch Newsletter","slug":"the-batch","description":"Weekly AI news and perspective for engineers, executives, and enthusiasts.","feature_image":null,"visibility":"public","og_image":"https://charonhub.deeplearning.ai/content/images/2021/08/thebatchlogotwitter-3.png","og_title":"The Batch Newsletter | DeepLearning.AI","og_description":"Weekly AI news for engineers, executives, and enthusiasts.","twitter_image":"https://charonhub.deeplearning.ai/content/images/2021/08/thebatchlogotwitter-2.png","twitter_title":"The Batch Newsletter | DeepLearning.AI","twitter_description":"Weekly AI news for engineers, executives, and enthusiasts.","meta_title":"The Batch Newsletter | DeepLearning.AI","meta_description":"Weekly AI news for engineers, executives, and enthusiasts.","codeinjection_head":null,"codeinjection_foot":null,"canonical_url":null,"accent_color":null,"url":"https://charonhub.deeplearning.ai/tag/the-batch/"},"url":"https://charonhub.deeplearning.ai/issue-326/","excerpt":"The Batch AI News and Insights: AI agents are getting better at looking at different types of data in businesses to spot patterns and create value. This is making data silos increasingly painful.","reading_time":14,"access":true,"comments":false,"og_image":null,"og_title":null,"og_description":null,"twitter_image":null,"twitter_title":null,"twitter_description":null,"meta_title":"OpenAI Reorgs For Profit, MiniMax-M2 Leads Open Coding, Universal Music Group Embraces AI, and more...","meta_description":"The Batch AI News and Insights: AI agents are getting better at looking at different types of data in businesses to spot patterns and create value...","email_subject":null,"frontmatter":null,"feature_image_alt":"Robots extract colorful data streams from silo towers, highlighting data silos being broken.","feature_image_caption":null},"seoImage":{"url":"https://charonhub.deeplearning.ai/content/images/2025/11/Tear-Down-Data-Silos--1.png","dimensions":{"width":1200,"height":676}},"banner":{"title":"BreakIntoAI with Machine Learning Specialization","databaseId":28989,"id":"cG9zdDoyODk4OQ==","featuredImage":{"node":{"altText":"Become an AI professional with one of the world's most popular Machine Learning courses. #BreakIntoAI with Machine Learning Specialization","mediaItemUrl":"https://home-wordpress.deeplearning.ai/wp-content/uploads/2023/03/4.png"}},"bannerCustomFields":{"bannerUrl":{"url":"https://www.deeplearning.ai/courses/machine-learning-specialization/","isUrlExternal":null}}},"announcementBanners":{"nodes":[{"title":"Jupyter C1","date":"2025-11-03T07:47:11","databaseId":37377,"announcementBannerCustomFields":{"text":"✨ New course! Enroll in","textColor":"#FFFFFF","courseLink":"https://bit.ly/48WUp7L","courseName":"Jupyter AI: AI Coding in Notebooks","backgroundColor":"#05256C","backgroundGradientColor":"linear-gradient(308.78deg, #F8A50F 0%, #8B0000 89.77%)","isOpenInNewTab":true}},{"title":"PyTorch","date":"2025-10-29T08:47:09","databaseId":37355,"announcementBannerCustomFields":{"text":"✨ New course! Enroll in the","textColor":"#FFFFFF","courseLink":"https://bit.ly/3X1M0Zr","courseName":"PyTorch for Deep Learning Professional Certificate","backgroundColor":"#05256C","backgroundGradientColor":"linear-gradient(104.95deg, #E94437 0.78%, #D43C3F 7.64%, #A02854 23.32%, #4E0875 43.9%, #37007F 49.78%, #0A003F 98.77%)","isOpenInNewTab":true}},{"title":"Post-training (AMD)","date":"2025-10-28T07:44:57","databaseId":37347,"announcementBannerCustomFields":{"text":"✨ New course! Enroll in","textColor":"#FFFFFF","courseLink":"https://bit.ly/4ofRjka","courseName":"Fine-tuning and Reinforcement Learning for LLMs: Intro to Post-Training","backgroundColor":"#05256C","backgroundGradientColor":"linear-gradient(106.06deg, #019966 0.78%, #028460 17.44%, #055153 51.73%, #0A003F 98.77%)","isOpenInNewTab":true}},{"title":"Databricks C2","date":"2025-10-22T07:32:37","databaseId":37315,"announcementBannerCustomFields":{"text":"✨ New course! Enroll in","textColor":"#FFFFFF","courseLink":"https://bit.ly/4qk9ZjY","courseName":"Governing Data Agents","backgroundColor":"#05256C","backgroundGradientColor":"linear-gradient(120deg,rgba(4, 53, 93, 1) 0%, rgba(255, 95, 70, 1) 100%)","isOpenInNewTab":true}},{"title":"Google C6","date":"2025-10-15T07:28:46","databaseId":37278,"announcementBannerCustomFields":{"text":"✨ New course! Enroll in","textColor":"#FFFFFF","courseLink":"https://bit.ly/491Hb9I","courseName":"Building Live Voice Agents with Google’s ADK","backgroundColor":"#05256C","backgroundGradientColor":"linear-gradient(105deg, #E94437 0.78%, #4385F3 50%)","isOpenInNewTab":true}}]}},"__N_SSG":true}

A MESSAGE FROM DEEPLEARNING.AI

News

OpenAI Reorganizes For Profit

Open-Weights Coding Leader

AI Music With Major-Label Support

Masking Private Data in Training Sets

Fish AI Reader

FishAI

联系邮箱 441953276@qq.com

相关标签