• Perle Labs Secures $17.5M in Funding
  • Season 1 is live!
  • Perle Labs Secures $17.5M in Funding
  • Season 1 is live!
  • Perle Labs Secures $17.5M in Funding
  • Join The Beta Launch
  • Perle Labs Secures $17.5M in Funding
  • Join The Beta Launch
  • Perle Labs Secures $17.5M in Funding
  • Join The Beta Launch
  • Perle Labs Secures $17.5M in Funding
  • Join The Beta Launch
New annotation quest shipped, we dropped a DEEP dive on data poisoning, & the community keeps growing

Welcome to The AI Loop #16. ETHDenver is officially in the rearview, and the team is back home and building. We launched new annotation quests, dropped a deep dive on data poisoning, and the community keeps growing.

Meanwhile, the AI world just had one of the wildest weeks in recent memory: record earnings, hundred-billion-dollar deals, and a software market shakeup that no one saw coming. Let's get into it.

Perle Labs Update

Post-ETHDenver: New Quests, New Research, Nonstop Community

We're back from an incredible month on the road: Korea, Hong Kong, Denver, New York, and the team has not slowed down. CEO Ahmed Rashad sat down with BeInCrypto for an interview on why sovereign data integrity is becoming essential for AI systems, especially as AI moves into high-stakes applications like medical diagnoses and autonomous vehicles.

Image
Perle Labs Pet Bounding Content Launched This Week!

On the product side, we launched the Pet Bounding Box Quest, a new annotation task where users review pet images and draw bounding boxes to train object detection models. The community says it's the most fun task we've launched so far. We also wrapped the final week of Human CAPTCHA: Round 7, our ongoing challenge where annotators identify real photographs among AI-generated images.

On the research front, we published a new article: "Adversarial Data Poisoning: How Bad Actors Corrupt AI Systems," diving into the epistemic breach from model collapse to active corruption. Frontier robotics teams are rebuilding data pipelines because generic licensed datasets aren't cutting it. The focus is shifting from "more data" to precise, domain-specific data. That's exactly what we're building.

The community continues to grow fast. 1 million annotators, 1 billion+ points distributed, and the Perle of Fortune winners were drawn live on Discord this week. We're just getting started.

Top AI Stories

Nvidia Posts Record $68B Quarter, Guides to $78B: AI Boom Is Alive and Well

Nvidia reported fiscal Q4 2026 earnings on Wednesday, crushing estimates across the board. Revenue hit $68.13 billion, up 73% year-over-year, with data center revenue alone reaching $62.3 billion. Net income nearly doubled to $43 billion. But the real headline was guidance: Nvidia projected Q1 revenue of approximately $78 billion, well above the $72.6 billion Wall Street consensus.

CEO Jensen Huang pointed to "skyrocketing" adoption of AI agents as the primary driver. The results arrived at a critical moment — markets had been rattled all week by fears that AI could disrupt legacy software companies faster than expected. Nvidia's numbers were a reminder that, whatever disruption AI causes downstream, the infrastructure buildout is only accelerating.

Source: Nvidia Newsroom

AMD and Meta Sign Landmark ~$100B AI Chip Deal

AMD and Meta unveiled one of the largest semiconductor partnerships in history on Tuesday. Under the multi-year, multi-generation agreement, Meta will deploy up to 6 gigawatts of AMD's Instinct GPUs — starting with custom MI450 chips — alongside AI-optimized EPYC CPUs across its data centers. First shipments begin in the second half of 2026.

The deal includes a performance-based warrant giving Meta the option to acquire up to 160 million shares of AMD, roughly a 10% stake. AMD CEO Lisa Su called it one of the "most transformative deals" in AMD's history, while Meta CEO Mark Zuckerberg said it allows Meta to "diversify our compute." The agreement validates AMD as a credible alternative to Nvidia in the AI chip race and signals that Big Tech's appetite for AI compute is far from peaking.

Source: CNBC

Anthropic Launches Enterprise Agents & Plugins, Triggers Massive Software Selloff

Anthropic made waves twice in a single week. On Tuesday, it launched Claude Cowork's enterprise agents program with pre-built plugins for finance, HR, legal, engineering, and design — plus new connectors for Gmail, DocuSign, FactSet, and more. Head of Product Matt Piccolella said: "We believe that the future of work means everybody having their own custom agent."

Separately, Anthropic published a blog post showing Claude Code can modernize legacy COBOL systems — a revelation that sent IBM stock plunging 13.2%, its worst single-day drop in 26 years. The broader software sector followed: Salesforce, Workday, CrowdStrike, and Datadog all fell sharply. It's no longer theoretical — AI agents are starting to reshape enterprise software in real time.

Source: TechCrunch

Stay in the Loop

That's a wrap on Edition #16. Record earnings, hundred-billion-dollar chip deals, and the enterprise software market getting a wake-up call. The AI world doesn't slow down, and neither do we.

Join the Perle Labs community on X, Discord, and Telegram to stay connected and catch the next drop of insights from across the ecosystem.

The AI Loop #16: Nvidia's $68B Quarter, a $100B AI Chip Pact, & Anthropic's Enterprise Agents Trigger Software Selloff