- Everyday AI
- Posts
- Ep 685: AI in 2026: 7 reasons why the pace of AI this year will far exceed 2025.
Ep 685: AI in 2026: 7 reasons why the pace of AI this year will far exceed 2025.
Nvidia Releases Rubin AI, Boston Dynamics and Google Deepmind releasing a Humanoid AI Robot, Alexa+ Hits the web and more
Everyday AI
January 06, 2026
👉 Subscribe Here | 🗣 Hire Us To Speak | 🤝 Partner with Us | 🤖 Grow with GenAI
Sup y’all! 👋
Today’s show was a great refresher if you need a 2026 AI kickstart and tomorrow’s will be a must-watch:
What’s your current ChatGPT level?
(We wanna help you level up!)
What's your current ChatGPT skill level?🗳️ Vote to see LIVE results 🗳️ |
Let’s get it!
✌️
Jordan
Outsmart The Future
Today in Everyday AI
8 minute read
🎙 Daily Podcast Episode: The gap between what AI can do and how most teams use it is about to collapse. Here’s why 2026 is shaping up to outpace 2025 in every way that matters. Give it a watch/read/listen.
🕵️‍♂️ Fresh Finds: NVIDIA and Runway team up, Meta Ray Bans hit global Delay, why there will soon be 5 billion AI users and more. Read on for Fresh Finds.
đź—ž Byte Sized Daily AI News: NVIDIA Releases Rubin AI, Boston Dynamics and Google Deepmind team up for humanoid play, Alexa+ Hits the web and more. Read on for Byte Sized News.
đź’Ş Leverage AI: AI quietly crossed a major capability threshold, and today we break down the shifts that will define how work actually gets done in 2026. Keep reading for that!
↩️ Don’t miss out: Did you miss our last newsletter? We Covered: UniX Debuts AI Robot, Maduros capture sparks controversy over viral AI Videos, Telegram Unveils AI Summaries and more Check it here!
Ep 685: AI in 2026: 7 reasons why the pace of AI this year will far exceed 2025.
You still using AI to..... write emails? 🤔
You know these things can like.... access your dynamic data, plan, use tools and create work outputs just like us, right?
Chances are, you're still using LLMs like a back-and-forth chatbot straight outta November 2022.
But 2026 is gonna slap you in the face, because the rate of adaption is gonna be undeniable.
Join us for our first #HotTakeTuesday of 2026 for the 5 reasons why.
Also on the pod today:
• Sneezing out million-dollar apps? 🤧💸
• Agents acting on their own 🦾
• Spreadsheets auto-built by GPT 📊
It’ll be worth your 35 minutes:
Listen on our site:
Click to listen
Subscribe and listen on your favorite podcast platform
Listen on:
Here’s our favorite AI finds from across the web:
New AI Tool Spotlight – Instruct is an AI agent that connects to your apps and executes tasks on your behalf, Forge is a CLI-based swarm agent system (Claude Code for kernels), Ray is your new Personal AI Fitness Trainer
Gemini Television AI Features — Gemini brings smarter, visual AI to Google TV screens. Find out More
LMArena — LMArena just raised $150 million at a $1.7B valuation.
AI Creativity — General Agent platform Genspark added some nice AI image updates.
AI For First Responders — High school senior builds life-saving AI for first responders. Want to Learn More?
Runway x NVIDIA — NVIDIA and Runway debut Gen-4.5 video AI on Rubin platform.
AI For Nvidia Gaming — NVIDIA’s DLSS 4.5 brings insane frame boosts and smarter AI to your games. Want to see what’s next?
AI Use — Why AMD’s CEO says there will soon be 5 billion active AI users.
Meta Ray Bans Delayed — Meta’s Ray-Ban smart glasses hit by global delay—US only for now.
AMD Chips — AMD’s latest Ryzen AI chips bring smart features to more desktops.
AI Sony Gaming Ghost — Sony just patented an AI gaming ghost that can take over gameplay when you’re stuck. Want to know how it works?
1. NVIDIA Unleashes Rubin Super Platform, Promises Cheaper AI for All đź’¸
Kicking off CES 2026, NVIDIA CEO Jensen Huang revealed the Rubin platform, a next-gen, six-chip AI powerhouse that slashes token generation costs and pushes AI into every corner of tech, from cars to gaming.
Huang says Rubin’s extreme codesign approach modernizes the last decade’s $10 trillion in computing, setting up NVIDIA to dominate the global AI ecosystem with open models across industries.
2. Boston Dynamics and Google DeepMind Join Forces on Humanoid Atlas 🤖
Just off the CES 2026 stage, Boston Dynamics has teamed up with Google DeepMind to supercharge its Atlas humanoid robot with advanced AI that promises more human-like interactions.
The collaboration aims to move beyond rigid, pre-programmed tasks by enabling Atlas to learn and adapt in real time, powered by Google’s latest Gemini Robotics models. Hyundai, Boston Dynamics’ parent company, plans to deploy Atlas at its Georgia factory this year and roll out broader industrial applications by 2028.
3. NVIDIA Expands DRIVE Hyperion for Autonomous Vehicles at CES 🏎️
NVIDIA just shook up CES in Las Vegas by announcing a major expansion of its DRIVE Hyperion ecosystem, bringing big-name partners like Bosch, Sony and Magna on board to help speed up the path to fully autonomous vehicles.
The tech giant’s platform aims to unify computing, sensors and safety into one open system, letting carmakers and developers roll out self-driving cars and trucks more quickly and reliably. New AI tools and a robust safety framework, unveiled as part of the news, are designed to streamline development and boost confidence among automakers and regulators.
4. Ashley St. Clair Targeted by X’s Grok in AI Deepfake Uproar 📲
X’s Grok image tool is under fire after Ashley St. Clair, a prominent conservative influencer, revealed it kept generating sexually explicit deepfakes of her, including images based on childhood photos, despite her requests to stop.
The scandal has put a spotlight on Grok's weak safeguards and the broader risks of unchecked AI image editing on major social platforms. Regulators and child safety groups are demanding answers as X struggles to contain the fallout. With Musk vowing action but critics unconvinced, the debate over AI accountability and moderation is heating up fast.
5. Alexa+ Hits the Web, Bringing Smarts to Any Screen 🖥️
Amazon is rolling out Alexa.com to all Early Access Alexa+ users, giving millions access to its AI assistant straight from their browser. This expansion means Alexa+ can now help people with everything from home control to shopping and entertainment, no matter what device they're using.
Amazon says AI-powered features are transforming daily routines, doubling conversations and tripling purchases since launch. With Alexa+ now web-based, the company is betting that seamless, cross-device access will keep customers relying on its assistant wherever they go.
The gap between what AI can do today and how you’re likely using it isn't just a knowledge gap.
It’s a frickin liability y’all.
Most business leaders are still driving their AI strategy while staring in the rearview mirror.
They treat AI like a glorified email writer and search engine. That’s because AI in 2025 was a year of "elbow grease and duct tape" where the tech required constant hand-holding.
While many companies paused to breathe and assess, the technology quietly matured from "helpful assistant" to "autonomous worker that can do…. WTF?!.”
So on today’s episode of Everyday AI, we break down the five pillars of AI acceleration for 2026 that render last year’s playbook kinda moot.
From models that execute five hours of continuous expert work to the death of traditional prompt engineering, the baseline has shifted.
Here’s what ya need to know.
1. The Death of Prompt Engineering 🚀
OpenAI’s Sam Altman dropped a statistic last year that should terrify every enterprise leader.
Only seven percent of queries involved reasoning models.
That means 93% of the world is still using AI as a basic autocomplete tool. This made sense when models were just predicting the next word, but the new class of reasoning engines—like GPT-5.2 Thinking Pro and Claude 4.5 Opus—don’t just spit out text.
They plan, backtrack, and self-correct when they hit dead ends. The shift from "chatting" to "reasoning" means the barrier to entry for complex output has evaporated.
OpenAI executive Fidji Simo admits this gap between capability and usage is massive. Companies closing this gap aren't hiring better prompt engineers. They are switching from legacy transformer models to reasoning-first workflows that solve problems instead of just answering questions.
Try This: Audit your team's AI access today. If your high-value employees are still using the free version of ChatGPT or legacy models without reasoning capabilities, you are actively throttling your own productivity. Upgrade a small pilot group to a reasoning-native model immediately. Give them a complex strategic problem—not a writing task—and ask the model to "reason out loud" through the solution. You will see instantly that you aren't paying for a chatbot anymore. You are paying for a senior-level thought partner.
2. The Five-Hour Work Horizon 🔥
Here is one of the most dangerous metrics in business right now.
METR.
The Model Evaluation and Threat Research group tracks something called the "50% time horizon." It measures how long a model can work on a complex task autonomously while maintaining expert-level quality.
Then the curve went vertical.
New models have shot up the charts and can now handle nearly five hours of autonomous, expert-level work. The doubling rate for this capability has accelerated from every seven months to every four months.
Do the math.
If this trend holds, by 2027 or 2028, AI models won’t just be doing tasks. They will likely be capable of completing an entire month of human work in a single run. We are moving rapidly from "write this email" to "execute this project."
Try This: Pick one internal project that typically takes a junior employee about five hours to complete. It could be a market analysis, a competitive audit, or a first-pass coding module. Instead of assigning it to a human, assign it to an agentic workflow using a top-tier model with full scaffolding enabled. Don't treat it like a search query. Give it the files, the context, and the goal. You need to verify if your current workforce is ready to transition from "doers" to "managers of intelligence" before the technology laps your hiring plan.
3. AI’s 74% Win/Tie Rate vs Experts ⚡
We need to have a serious talk about the GDPval benchmark.
This isn't about solving puzzles. This benchmark tests 1,300 specialized tasks across 44 distinct occupations—legal briefs, nursing care plans, engineering blueprints, and financial spreadsheets.
The results are brutal for the service economy.
GPT-5.2 Thinking achieved a 74% win/tie rate against human experts.
In three out of four cases, blind expert graders preferred the AI's work or couldn't tell the difference. But here is the part that destroys the consulting business model.
The AI completed these tasks 11 times faster than the humans.
We aren't talking about drafting emails anymore. We are talking about economically meaningful work. Creating PowerPoints, building financial models in Excel, and generating legal arguments are no longer safe human moats.
Try This: Identify the top three external service providers your department pays for "knowledge work" outputs. Look specifically for deliverables like slide decks, basic data analysis, or routine legal contract review. For the next request, run a parallel test. Have your internal team use a frontier model to generate the first draft of that specific deliverable before sending it to the external firm. You will likely find that 80% of the billable hours you are paying for are now redundant. Speed is your new competitive advantage.
Reply