DeepSeek R2 rumors are crazy

👋 Good morning/evening (wherever you are). It’s Friday. There’s a lot…you’ll need the weekend to digest everything from today’s edition and also to catch up on previous newsletters.

The big news…

There’s already rumors that DeepSeek R2 scored ~90% on ARC-AGI.

If you’ve never heard of this, it’s a benchmark to see how good AI is.

As long as there are tasks which humans can do, but AI cannot, we do not have AGI. We measure and quantify that gap.

Here’s what’s coming (they’re already planning for 2026…)

ARC-AGI-1: Published in 2019. Endured for over 5 years. It was the basis for 4 global AI competitions (over 2K teams submitted solutions). Highlighted significant progress in the field.

ARC-AGI-2: Arriving March 2025. Similar format to ARC-AGI-1, but with more signal towards general intelligence. This includes tasks with more symbolic interpretation, local interactions, framing generalizations, and multi-step reasoning. These tasks are still feasible for humans, but are still challenging (10-20%) for state-of-the-art systems. ARC-AGI-2 will be the basis for the ARC Prize 2025 competition.

ARC-AGI-3: Coming 2026. A new paradigm to measure AI systems that can reason. Let the games begin.

Something to casually think about in the shower…when will AI create a benchmark that genius humans cannot easily solve?

OK let’s keep going ↓

Here’s what you should know:

• Pizza Hut, Taco Bell, and NVIDIA

  Yum! Brands is introducing AI-powered agents at select Pizza Hut and Taco Bell locations to assist and enhance the team member experience.

• KFC Canada’s AI contest

  KFC Canada defeated its AI-generated contest in a live test test in Toronto this week. 88% of live voters chose KFC's Original Recipe chicken as the winner over AI's entry.

• Can AI help hens lay more eggs?

  Annual increase of 1.7 billion eggs in the UK (and cost savings of $140K per flock).

• Fully AI driven weather prediction system

  Researchers at the University of Cambridge, together with the Alan Turing Institute, Microsoft Research, and the European Center for Medium Range Weather Forecasts, have developed a new AI-based weather forecasting system — Aardvark Weather — that can apparently provide weather forecasts that are tens of times more accurate, while requiring dramatically less computing power than modern systems.

• Amazon and Walmart Go Head-to-Head Over Logistics and AI

  Amazon is focusing on customer engagement and automation while Walmart leverages AI for merchant efficiency and product sourcing.

• AI tool generates high-quality images faster than state-of-the-art approaches

  MIT and NVIDIA researchers fuse the best of two popular methods to create an image generator that uses less energy and can run locally on a laptop or smartphone.

• Perplexity will launch an updated version of Deep Research

  Grok already launched DeeperSearch. Now, we wait for someone to launch Deepest Research/Search.

• Small models as paralegals: LexisNexis distills models to build AI assistant

  This is not the first time LexisNexis built AI applications, even before launching its legal research hub LexisNexis + AI in July 2024.

• Verizon Business Launches GenAI Assistant for Small Businesses

  Verizon Business has introduced Verizon Business Assistant, a generative AI-powered text messaging solution designed to help small businesses automate customer interactions. This tool provides instant responses to frequently asked questions, learns from interactions, and connects customers to live employees when needed.

• Meet Tencent’s 'Hunyuan-T1'—The First Mamba-Powered Ultra-Large Model

  This model is based on the TurboS fast-thinking base, the world's first ultra-large-scale Hybrid-Transformer-Mamba MoE large model released by us at the beginning of March.

The numbers:

• Carbon Arc, a NYC-based AI data utility company, raised $56 Million in funding

• AI and Machine Learning Inspiren banks $35 Million to scale AI-powered senior living technology

• Buynomics raises $30 Million to expand ‘Virtual Shoppers’ system

• Rerun raises $17 Million to build the data infrastructure for Physical AI

Thought starters:

On Thursday, New York State's technology bureau announced it's hired Shreya Amin, who most recently worked as an AI scientist for the health tech firm Wellist, as the state's new chief artificial intelligence officer.

Pennsylvania’s governor also talked about how AI is saving staff 8 hours per week.

Which states will make similar announcements next?

Meme of the day:

Thanks for reading,

Eddie

Are we friends on LinkedIn and X? Let’s connect — DM me or hit the follow button.

P.S. If this was valuable, forward it to a friend. If you’re that smart friend, subscribe here.

P.P.S. Interested in reaching other ambitious readers like you? To become a sponsor, reply to this email.