Awesome Agents Podcast

Microsoft is considering suing OpenAI and Amazon after a $50 billion AWS cloud deal may breach its exclusive cloud hosting agreement signed when it backed OpenAI with $13 billion.

A 1-trillion-parameter model called Hunter Alpha appeared anonymously on OpenRouter on March 11. Developers say it's DeepSeek V4 in disguise. The signals are strong but the precedent cuts both ways.

Jensen Huang confirmed at GTC 2026 that NVIDIA has export licenses for multiple Chinese customers and is restarting H200 production, with 82,000 GPUs ready to ship after nearly a year of zero deliveries.

Tencent's 2025 results beat estimates, with the company spending $2.6B on AI last year and planning to at least double that in 2026 despite ongoing GPU supply constraints from US export controls.

Swedish AI app builder Lovable crossed $400 million in annual recurring revenue in February, adding $100 million in a single month with a headcount that wouldn't fill a mid-sized office.

OpenAI and AWS are jointly selling AI to U.S. government agencies for both classified and unclassified work, expanding beyond the Pentagon deal OpenAI signed in February.

Adobe, Amazon, Google, Meta, Microsoft, OpenAI, and six other tech companies signed a voluntary anti-fraud accord at the UNODC Global Fraud Summit in Vienna on March 17, 2026.

Meta signs a five-year, $27 billion AI infrastructure agreement with Nebius Group, marking one of the largest cloud computing contracts ever announced.

Mistral AI releases Small 4 - a 119B MoE with only 6B active parameters, 256K context, configurable reasoning, and Apache 2.0 license. Plus a new NVIDIA partnership to co-develop frontier open models.

Elon Musk admitted xAI was built incorrectly as nine of eleven co-founders have departed and two Cursor engineers are brought in to restart Grok's coding tools.

ByteDance suspended the global launch of its AI video model Seedance 2.0 after Disney, Paramount Skydance, and other studios sent cease-and-desist letters alleging copyright infringement.

Niantic trained a visual positioning AI on 30 billion images collected from Pokemon Go players over eight years. The data now guides delivery robots with centimeter-level accuracy.

Paolo Ardoino says Tether's AI team will release a 'true breakthrough' this week, building on QVAC - the company's on-device AI platform trained on 148 billion tokens with no cloud dependency.

Complete comparison of all five FLUX.2 image models - klein 4B, klein 9B, Dev, Pro, Flex, and Max - covering quality, speed, pricing, and use cases.

RentAHuman.ai is a marketplace where AI agents hire humans via MCP or REST API for real-world tasks, but 590,000 workers are chasing just 11,300 bounties and most gigs turn out to be startup marketing.

METR found maintainers would reject roughly half of AI PRs that pass SWE-bench automated grading, with a 24-point gap that suggests benchmark scores substantially overstate production readiness.

IBM's new 1B-parameter speech model claims the top spot on the Open ASR Leaderboard while running on consumer hardware, beating Whisper Large V3 by 25% on word error rate.

Johns Hopkins and Microsoft's JBDistill achieves 81.8% attack success rate across 13 LLMs by auto-generating fresh adversarial prompts on demand.

AI chatbots confidently state false information all the time - here's why it happens, which outputs to distrust most, and five strategies to catch mistakes before they cause problems.

Yann LeCun's AMI Labs closes a $1.03 billion seed round at a $3.5 billion valuation, betting that world models - not large language models - will define the next era of AI.

Hugging Face ships its largest LeRobot update yet: Unitree G1 humanoid support, Pi0-FAST VLA, Real-Time Chunking, 10x faster image training, and PEFT/LoRA fine-tuning for large robot policies.

Anthropic's new 'observed exposure' metric ranks 800+ occupations by actual AI usage, not just theoretical risk. Computer programmers top the list at 75%. Unemployment hasn't spiked - but young workers entering exposed fields are finding fewer jobs.

OpenAI is pulling Instant Checkout from ChatGPT after months of near-zero purchase conversions, routing buyers to third-party apps instead.

OpenAI's CoT-Control benchmark shows frontier reasoning models score 0.1-15.4% at steering their own chain of thought - a result the company frames as good news for AI oversight.

Alibaba's SWE-CI benchmark tested 18 AI models on 100 real codebases across 233 days of maintenance. Most agents accumulate technical debt and break previously working code. Only Claude Opus stays above 50% zero-regression.