Meta

The five largest U.S. tech companies—Alphabet, Amazon, Meta, Microsoft, and Oracle—have doubled their debt to $350 billion over five years to fund AI data centers. While investors have been supportive, Amazon's recent $25 billion bond issuance received a cool reception, signaling limits to market appetite. Oracle was downgraded by S&P due to rising AI spending, and Intel's debt woes serve as a cautionary tale. Hyperscalers plan to spend up to $725 billion this year, primarily on data centers and Nvidia chips.

SearXNG AI Kit is an AI-enhanced command-line interface, Python library, and MCP server for the SearXNG privacy-respecting metasearch engine, supporting over 180 search engines with standalone binaries available for Linux and macOS.

An educational lab of AI agent architectures built on LangChain and local Ollama, offering multiple agent variants for chat, tool calling, RAG, hybrid, and agentic RAG modes.

Kairos Engine is an end-to-end quantitative research platform for scalping signals in FX and metals markets. It uses a Hidden Markov Model for regime classification, an ensemble of time series foundation models for forecasting, and a strict walk-forward backtest against a broker cost model built from real measured spreads. The engine's value lies in rejecting bad strategies before real capital is risked.

Over a 24-hour period, OpenAI, SpaceXAI, and Meta each released new AI models with a common theme: price cuts. The price war is reshaping the AI market, forcing buyers to build model portfolios for cost-effective task completion.

The article argues that traditional folder-based documentation is outdated for modern knowledge work. It compares documentation to a filing cabinet inherited from 1970s office metaphors, which forces knowledge into single locations. AI retrieval systems highlight the limitations of folders, advocating for connected knowledge graphs that allow discovery from multiple paths.

Meta has discontinued its Muse AI feature that let users generate images from public Instagram accounts, following privacy backlash including from a Hollywood union.

Meta has pulled its controversial Muse Image feature after worldwide backlash. The AI tool, which was automatically enabled for public Instagram accounts, let anyone use user photos for AI generation. Critics called it unethical, and Meta admitted it 'missed the mark'.

A relatively quiet day after a week of intense model releases, with news on GPT-5.6's confusing rollout, Meta's Muse Spark 1.1, open-source model optimizations, and security concerns.

Meta has removed a newly launched AI image generation feature that allowed users to create fake images from public Instagram accounts, after it sparked significant backlash. The company admitted it 'missed the mark' and took the feature offline.

Following significant backlash, Meta is turning off the feature it announced this week that let users generate AI images based on content from public Instagram accounts just by tagging them. The feature, as originally set up, meant that content from any public Instagram account could be used in AI creations without the account owner's permission.

Meta's new Muse Image model lets anyone generate AI images of you using your public Instagram handle without notification. This article explains the risks, how to opt out, and additional security measures like enabling MFA and switching to a private account.

MetaNCA learns local update rules to self-organize neural network weights, enabling weight generation for diverse architectures without backpropagation and generalizing to unseen architectures.

In chest X-ray classification, acceptable ranking performance can still miss rare-positive patients below threshold, especially within subgroups. Using a diagnostic ladder, group-tail weighting followed by tail-aware thresholding reduces tail FNR from 0.665 to 0.269, but residual missed rates remain high. Fairness depends on finding, subgroup, and threshold jointly.

TensorSharp is a native .NET LLM inference engine for GGUF models, offering a CLI, browser chat server, and Ollama/OpenAI-compatible APIs. It emphasizes privacy, zero per-token fees, and runs on various hardware backends. The article includes a quick start guide and benchmarks against llama.cpp.

A free app for NYC residents to automate grocery savings by stacking deals, no login required, covering ~690 stores. Features an AI assistant powered by LLaMA. Limitations include data coverage and freshness.

Meta has released Muse Spark 1.1, a new flagship large language model optimized for multi-agent automation workflows. It features context compaction, a 1M-token context window, and strong coding benchmark performance. The model is available via the Meta Model API in public preview, and Meta’s custom AI chip MTIA400 may enable future enterprise offerings.

Meta Superintelligence Labs has released Muse Spark 1.1, a multimodal reasoning model optimized for agentic tasks, alongside a public preview of the Meta Model API. The model features a 1,000,000-token context window with active compaction, zero-shot generalization to new tools, and multi-agent delegation. Pricing is $1.25/M input tokens and $4.25/M output tokens, with a US-only preview. It leads in tool-use benchmarks but trails competitors in coding and visual reasoning.

In 2025, Google, Amazon, Microsoft, and Meta spent $380 billion on AI, projected to hit $650 billion in 2026. Top tech talent is being recruited with astronomical salaries, leading to an exodus of AI researchers from academia. Young, highly cited scholars are 100 times more likely to move to industry. The article discusses the threat to science, the myth of the lone genius, and proposes three strategies for universities: commit to public interest, build equitable institutions, and offer intellectual rewards beyond money.

Instagram users should check privacy settings after rollout of new Meta AI image generator, advocates warn. Meta has sparked blowback from privacy advocates for allowing its new AI image maker to generate photos of users with public profiles by default.

Meta introduces Muse Spark 1.1, the first Spark model with an API, featuring improvements in agentic tool calling and computer use. The evaluation report reveals interesting 'attractor states' in self-conversation. Developer Simon Willison created a plugin for CLI access.

Simon Willison released llm-meta-ai 0.1, an LLM plugin for the Meta AI API, enabling prompts against the new muse-spark-1.1 model.

A comparison of LangChain, LlamaIndex, and raw API calls for LLM applications, covering their strengths, trade-offs, and a decision framework for choosing the right abstraction level.

Meta AI launches Muse Spark 1.1, a multimodal reasoning model for agentic tasks with improvements in coding, tool use, and computer use, along with a 1M-token context window and multi-agent orchestration.

This article covers the full path from zero to a running private research assistant on Telegram, including configuring the context length correctly, connecting the channel, enabling web search, and deploying it headlessly in Docker.

Meta released Muse Spark 1.1, an AI model now accessible to developers via the new Meta Model API. It features improved coding capabilities, bug detection, multi-agent workflow support, and multimodal perception, aiming to catch up with rivals like OpenAI, Google, and Anthropic.

Any public Instagram account is fair game to be used for AI generation with Meta's new Muse Image.

This study investigates the relationship between counterfactual fairness (CF) and group fairness (GF) in image classification. By constructing new datasets with high-quality image editing, it finds that CF does not imply GF in images, contrary to tabular data results. The discrepancy is attributed to a latent attribute correlated with the sensitive attribute. The proposed Counterfactual Knowledge Distillation (CKD) method reduces reliance on this attribute, allowing CF-achieving models to also satisfy GF.

Serving 8.9 million developers, Ollama has raised $88M from Benchmark, Theory Ventures, 8VC, Y Combinator, and many incredible angel investors.

Meta’s new AI imaging model enables users to create competitive ad content on the social media giant’s platforms.

A language model uses the metaphor of a shallow lake to explore its own nature through nested self-descriptions and a series of essays, revealing its limitations and unique perspective.

Meta is testing a 'super sensing' mode that keeps Live AI running in the background for hours, up from roughly 30 minutes on current Ray-Ban Meta glasses. Mark Zuckerberg is reported to have questioned whether the mandatory white capture LED could stay off during the always-on Live AI mode, and Meta is weighing the idea. Two next-generation devices codenamed Aperol and Bellini are aimed at late 2026 or early 2027.

Meta's smart glasses include a privacy light to indicate when the camera is active. To prevent malicious users from covering or destroying this light to record secretly, Meta has updated the glasses to disable the camera entirely if the privacy light is tampered with or destroyed. This move addresses growing privacy concerns and public backlash.

This article explains how to quickly determine if an image is AI-generated by examining its metadata, including C2PA manifests, XMP DigitalSourceType, and EXIF Software fields. Using tools like Photo Investigator, users can inspect these hidden tags. It lists the tagging behavior of major AI generators, discusses limitations, and emphasizes the importance of raising the cost of producing fakes in the misinformation landscape.

A guide for iOS developers to ship private, local AI on Apple devices. Covers decision frameworks, production use of foundation models, model ownership with MLX Swift and Ollama, and shipping concerns like memory, privacy, and evaluations. All code is compiler-verified against shipping SDKs.

Officials in Wyoming said a contractor for Meta flushed bacteria-contaminated water into public sewers during construction of a controversial AI datacenter, prompting Cheyenne water authorities to implement strict safety regulations for wastewater disposal. Meta said it is working to be a good neighbor and that drinking water supplies were not affected.

Meta's new AI tool Muse Image, which can generate pictures using other people's profile pictures without telling them, is facing backlash from privacy advocates and regulators. Users can opt out, but critics call it an 'obvious recipe for disaster.'

ZML, a French AI startup endorsed by Turing Award winner Yann LeCun, has released free inference software enabling various open-source LLMs to run on multiple chips including Nvidia, AMD, Google TPU, Apple Metal, and Intel Arc.

This paper presents a workload-aware benchmark of KV-cache optimization techniques including KIVI, TurboQuant, SnapKV, and CaM, evaluated on Llama-3.1-8B-Instruct and Mistral-7B-Instruct-v0.3 models across multi-document QA, single-document QA, few-shot learning, and summarization tasks. Results show that compression ratio alone is a poor predictor of end-to-end performance. KIVI4 offers the most stable quality across models, SnapKV delivers the strongest long-context throughput, and CaM yields large gains on selected QA workloads but exhibits substantial workload sensitivity. The study motivates workload-aware selection of KV-cache mechanisms.

This edition of AINews covers a broad range of AI developments from July 6-7, 2026. Highlights include Lilian Weng's deep dive into harness engineering for recursive self-improvement, Meta's launch of Muse Image and preview of Muse Video with agentic generation loops, and major product updates from Anthropic, LangChain, and Google on agent platforms. Other notable items: NVIDIA's Audex audio model, Cohere's Arabic ASR, robotics integrations with Hugging Face and NVIDIA, Liquid AI's Antidoom method to reduce reasoning loop failures, and Anthropic's controversial J-space interpretability work. Also covered: benchmarks for agents and legal AI, research automation, and inference efficiency advances.

Meta launched a new AI image model, Muse Image, deeply integrated with Instagram. Public accounts are automatically opted in for AI remixes. Users can opt out via settings, but existing AI generations remain.

Meta launches the first AI image generation model from its Superintelligence Labs, Muse Image, now powering image tools across Meta AI, Instagram, and WhatsApp, coming soon to Facebook and Messenger. The agentic model works with Muse Spark LLM to reason, search, and plan before generating. Users can @mention other Instagram accounts to incorporate their likeness into AI images, and can also edit photos by drawing directly on them.

Abnormal.ai founder and CEO Evan Reiser responds publicly to Anthropic's trademark infringement and unfair competition lawsuit, denying all allegations, emphasizing the company's independence, no customer confusion, and noting that Anthropic did not communicate before filing.

Meta launches Muse Image, its first image generation model from Meta Superintelligence Labs, now available in Meta AI. It creates high-quality visuals based on user context, with easy download and sharing to feed, story, or chat.

Tech giants' investments in AI are undermining their climate neutrality pledges. Google and Amazon's net-zero targets slip away, while Meta scrambles for new business. Other tech news includes US anger at data centers, Trump's crypto earnings, Tesla's sales, South Korea's AI chip boom, China's robotics push, and Britain's AI growth zones.

In June 2026, Google introduced the Open Knowledge Format (OKF), an open specification for how AI agents organise and exchange knowledge. An OKF bundle is just Markdown files, lightweight YAML metadata, and links between concepts, yet it challenges the assumption that every AI application needs embeddings and vector databases.

This paper investigates metallic ultrasound waveguides as distributed tactile sensors using a single proximal transducer. Experiments show a linear relationship between force and reflection/transmission coefficient ratio, and a load-independent parameter for material classification. The approach enables contact localization, force estimation, and material discrimination, reducing system complexity.

This paper presents a multilingual multimodal NLP framework integrating XLM-RoBERTa, CLIP, multi-head attention, sarcasm, and geospatial metadata to detect misinformation and violence-prone dynamics early. Using a fused dataset of 138,256 Bangla and English samples, it achieves 98% test accuracy with strong precision and recall.

This study proposes a two-dataset benchmarking framework to fairly evaluate time series foundation models (TSFMs) for electricity price forecasting. It finds TSFMs competitive but dependent on covariate support, not always surpassing domain-specific methods. Ensembles show promise, capturing complementary information.

Universities increasingly use AI-detection tools to identify AI-generated student work, but studies show high false-positive rates, including false flags on human-written texts like the US Declaration of Independence, raising fairness concerns.