AI Briefing
TLDR AI news
The latest AI and developer news from TLDR AI, with short summaries.
Visit website →TLDR Seeks First Dedicated Product Manager for Applied AI Team
TLDR is hiring its first dedicated Product Manager for its Applied AI department, offering $180k-$225k base salary plus up to $60k bonus.
OpenAI Rolls Out Ultrafast Mode for GPT-6.1 Sol in API, Codex, and ChatGPT Work
The new mode offers near-Astra intelligence at up to 8x faster speeds than Sol Standard, priced at $12 per million input tokens.
Midjourney Tests Thinking Mode for Image Generation on Alpha Platform
The new mode aims to improve prompt accuracy, typography, and coherence in generated images.
OpenAI's annualized revenue reportedly $20 billion lower than previous projections
The Financial Times reports OpenAI told investors its annualized revenue is approaching $50 billion, down from earlier reports of nearly $70 billion.
Wake-sleep Algorithm Boosts Long-Horizon Legal Agent Performance to 15.7% All-Pass Rate
Applying the wake-sleep algorithm to long-horizon legal agents increased the all-pass rate from 2.9% to 15.7% and halved the cost per task through selective memory retrieval.
Exa Previews ATLAS Benchmark to Evaluate Agentic Search Accuracy and Cost
ATLAS uses 547 tasks with a $0.24 grading cost to reveal that no agent achieves a row F1 score above 0.5 under a $1 budget.
Speculative Decoding Speeds Up LLMs by Shifting GPU Work from Memory-Bound to Compute-Bound
Speculative decoding accelerates LLM inference by increasing total FLOPs to utilize idle compute resources, shifting the workload from memory-bound to compute-bound at low batch sizes.
Epoch AI: Frontier Models Excel at Defined Tasks but Fail at Open-Ended Work Automation
Epoch AI's evaluation of six frontier models reveals that while closed-weight models like Claude Fable 5.1 and GPT-6 Astra excel at defined coding and computational tasks, they lack the judgment for open-ended work, often misinterpreting experimental flaws and failing to adhere to implicit standards.
Ex-OpenAI, Ramp, and Cognition Staffers Launch Hone with $60M to Build Autonomous Business Agents
The startup, valued at $285 million, is backed by Benchmark and Index Ventures and already counts Cognition and Modal as early customers.
Scale AI Introduces HSS Benchmark to Test Intuitive Visual Reasoning in Multimodal Models
Human participants achieve 93.1% accuracy on the new HSS benchmark, while the top-performing model, GPT-6-astra, reaches only 53.6%.
LlamaIndex launches OpenDocRouter to unify document parsing across frontier and open-source models
The new platform offers token-based billing with $5 in free credits for new accounts and supports synchronous parsing for documents up to 50 pages.
Analysis of Anthropic's Pre-IPO Corporate Structure and Board Control
The Long Term Benefit Trust holds four of seven board seats and controls key governance mechanisms, though its trust agreement remains non-public.
Perplexity Releases pplx-embed-v2-late: Multi-Vector Embeddings for Text and Visual Documents
Perplexity has released pplx-embed-v2-late, a family of ColBERT-style multi-vector embedding models in 0.6B and 9B sizes, supporting multimodal retrieval for text and visual documents.
Oracle, Broadcom, and SpaceX Seek Blockbuster Debt Deals to Pay for AI Chips
Broadcom is arranging more than $50 billion in financing for OpenAI's custom AI chip.
Musk’s Grok Bot to Use Rival Models Including Claude Opus 5.5, Midjourney, and Suno
Elon Musk announced that Grok Bot, the AI agent from SpaceXAI, will now utilize a multi-model backend including rival APIs like Claude Opus 5.5, Midjourney, and Suno.
Tony Fadell: First Wave of AI Gadgets Failed by Ignoring Real User Needs
The iPod creator argues that successful AI agents must operate on-device to earn consumer trust, citing Rabbit R1 and Humane Ai Pin as cautionary tales.
Texas ERCOT Pauses Switch-On for Large Data Centers Pending Batch Zero Audit
ERCOT has paused approvals for data centers of 75 MW or more to switch on, affecting 17 projects, pending a December audit of the Batch Zero process.
Project Zero outlines emergency patching systems for urgent vulnerabilities
The guide highlights feature flags, filtering, and hotpatching as key methods to bypass slow standard update cycles.
Meta and Sierra Launch Personal Agent Protocol for Secure AI-Brand Interaction
Meta and Sierra, along with industry partners, have announced the Personal Agent Protocol, an open standard enabling personal AI agents to securely interact with businesses via websites, APIs, or company agents using OAuth.
Hark debuts Hark Pro AI agent ahead of 2027 hardware launch
Hark has launched Hark Pro, an AI agent available on web and mobile with free, $20, and $100 monthly tiers, preceding its hardware devices scheduled for 2027.