- Everyday AI
- Posts
- Ep 698: Human-AI Collaboration: Best practices for working alongside AI
Ep 698: Human-AI Collaboration: Best practices for working alongside AI
Best practices for human-AI collab, Google brings Personal Intelligence into AI Mode, DeepMind strikes a deal with Hume AI, and OpenAI wants a piece of client revenue?
Everyday AI
January 23, 2026
👉 Subscribe Here | 🗣 Hire Us To Speak | 🤝 Partner with Us | 🤖 Grow with GenAI
Outsmart The Future
Today in Everyday AI
8 minute read
🎙 Daily Podcast Episode: AI assistants are everywhere, but working with them is still harder than it should be. The real shift now isn’t better prompts — it’s learning how humans and AI actually collaborate. Give it a watch/read/listen.
🕵️‍♂️ Fresh Finds: Waymo launches in Miami, Krea AI adds real-time image editing, Runway’s test shows 90% couldn’t tell AI video from real footage and more Read on for Fresh Finds.
🗞 Byte Sized Daily AI News: Google brings AI Personal Intelligence into Search, DeepMind strikes a deal with Hume AI and brings on CEO Alan Cowen, and DeepMind’s CEO calls OpenAI’s ad rollout early and risky and more Read on for Byte Sized News.
💪 Leverage AI: AI didn’t just get smarter. It changed the job. The new advantage isn’t knowing how to talk to models — it’s knowing how to manage them. Keep reading for that!
↩️ Don’t miss out: Did you miss our last newsletter? We Covered: Meta's new AI models are (kinda) out, Apple's plan for AI pin revealed, Microsoft releases new vision model for robotics and more. Check it here!
Ep 698: Human-AI Collaboration: Best practices for working alongside AI
Spending more time fixing your AI outputs then you're saving?
You're not alone.
The trap?
You're in operator mode. Falling for the industry status quo like upskilling and human-in-the-loop.
The real winners in the AI race?
Companies that have changed the human-AI relationship.
How?
Join us for Volume 4 of our Start Here Series as we uncover what you need to know.
Human-AI Collaboration: Best practices for working alongside AI -- An Everyday AI Chat with Jordan Wilson
Also on the pod today:
• Operator to orchestrator mindset shift đź§
• "Human in the loop" is outdated 🚫
• Upskilling vs. unlearning debate 🔄
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 – Tonkotsu Manages an entire team of coding agents from a doc, Krisp is an AI Meeting Assistant with #1 Noise Cancellation, World API is a public interface for generating explorable 3D worlds using World Labs’ multimodal world model, Marble.
Spotify AI Playlist — Control your playlists with plain language—Spotify generates personalized mixes from your prompt.
Waymo in Miami — Waymo opens Miami to paying robotaxi riders, expanding its U.S. lead
Realtime Edit Beta — Realtime image edits, now live — expect instant, complex changes.
Runway AI Test — Over 90% couldn't tell AI videos from real — can you?
AI Slides Creative Mode — Genspark just added a one-click "Redesign Layout" for slides, Try it now to see new layouts without manual tweaks.
Gemini for Workspace Business — Google Workspace finally gets Gemini ultra — major AI boost for businesses.
AI Sleep Study — One night of sleep data can flag risks for 100+ diseases. Curious which conditions your sleep might be hiding?
Cursor 2.4 Release — Cursor 2.4: agents spawn subagents and generate images inline—faster, parallel edits.
ChatGPT Instant Personality Update — OpenAI tweaked GPT-5.2 Instant to sound more conversational. Want to see if it actually sticks?
1. Google adds Personal Intelligence to AI Mode in Search 🔍
Google announced today that Personal Intelligence is rolling out to eligible AI Pro and Ultra Search subscribers in the U.S., letting users opt in to connect Gmail and Google Photos for personalized Search results.
The change means Search can now use your personal context to tailor recommendations and insights, while Google says connections are optional, privacy protections limit direct training on inboxes and photos, and feedback tools are available. The feature is launching as a Labs experiment for personal accounts only and will appear via an invitation or in Search settings for eligible users.
2. Senator Markey presses AI firms over ChatGPT ads 🧑‍💻
Sen. Ed Markey is demanding answers now that OpenAI plans to test ads in free ChatGPT, warning that embedding ads in conversational AI raises new consumer protection, privacy, and child-safety concerns.
In letters to OpenAI and six other major AI and tech companies, he questions whether chatbots will use sensitive user queries to target ads later, and whether users can reliably tell sponsored content from genuine responses. OpenAI says ads won’t appear for under-18s or during health or political conversations, but Markey calls that promise insufficient and warns the emotional bond users form with chatbots could be exploited.
3. Google DeepMind scoops Hume AI leaders in license-backed acqui-hire 🤝
According to Wired and TechCrunch, Google DeepMind has licensed technology from voice startup Hume AI and hired CEO Alan Cowen plus about seven top engineers to boost Gemini’s voice and emotional-audio features, while Hume retains a non‑exclusive rights plan and will keep selling its models.
The move, structured as a talent acquisition rather than a full buyout, lets DeepMind quickly strengthen conversational and emotion-aware voice capabilities without absorbing the whole company. Regulators are watching these team-focused deals more closely because they can concentrate talent and capabilities without triggering traditional merger scrutiny.
4. OpenAI to start sharing revenue from AI-driven discoveries đź’¸
According to a detailed report in The Information, OpenAI is shifting its 2026 strategy from pure subscription and API fees to outcome-based licensing and IP deals that let the company take a cut of value created by AI in high-value fields like drug discovery and scientific research.
The move signals a major business-model change: OpenAI wants payment tied to real-world results rather than just usage, aiming to capture long-term revenue as AI contributes to discoveries. That strategy raises privacy, regulatory, and market-power questions because outcome-based splits and IP arrangements could reshape who benefits from research and how data is used.
5. Hassabis says OpenAI’s ad move is early and risky⚠️
Google DeepMind CEO Demis Hassabis told Axios at Davos he was “a little bit surprised” that OpenAI has already started testing ads inside its chatbot, noting the move raises questions about trust and the assistant relationship.
He warned ads can undermine the helper role of chatbots, arguing conversational assistants differ from search and require more careful thought before monetization is added. Hassabis said Google currently has no plans to insert ads into its AI chatbot and emphasized a cautious, scientific approach as Google expands personalization features for Gemini.
6. OpenAI’s API hits $1B ARR in a single month 💰
OpenAI says its API business added more than $1 billion in annual recurring revenue in the past month, underscoring that the company’s engine for powering other firms is now a major revenue driver.
The revelation, posted by CEO Sam Altman, arrives as OpenAI pivots beyond consumer subscriptions to offset rising compute and data-center costs and explores new monetization like ads and licensing. That shift signals the company is treating enterprise integrations and downstream licensing as strategic growth levers rather than relying solely on ChatGPT subscriptions.
Your best AI habits are now your biggest liability.
Check that shorty. Not your WORST habits. Your BEST ones.
The prompting skills that made you dangerous in 2024? The iteration loops you perfected? They're actively holding you back in 2026.
Because the skill gap shifted. It's not technical anymore. It's managerial. And the teams pulling ahead aren't better at injecting AI on top of tried-and-true workflows.
They’ve rebuilt workflows for AI-native orchestration.
That's why we kicked off volume four of our Start Here Series this week. We're tackling the human-AI collaboration gap that's quietly separating winners from everyone else.
Here's the brutal truth: if you're still operating AI, you've already lost. The new game is orchestrating it. And that requires killing some habits you worked hard to build.
1. The old safety net is broken đź’€
You've been told humans catch AI's mistakes. That someone on your team is the safety net.
That worked when AI moved slow.
Not anymore.
Agentic AI creates action traces so long and so fast that no human can realistically interpret them. And when AI is right 85 to 95% of the time? Your "safety net" falls asleep.
Over 700 court cases worldwide now involve AI hallucinations. That number climbs daily.
Generic oversight isn't oversight. It's a comforting lie. And it's the first habit you need to kill.
Try This:
Be honest about their expertise in that specific domain.
If you're putting generalists on expert-level work, you're running on hope.
This week, swap one reviewer for someone who knows the domain cold and track the difference.
2. Operators are out, orchestrators win 🎯
Here's what the shift actually looks like.
Your job isn't doing the work anymore. It's defining what the work looks like, what information it needs, and what success means. The AI does the work. You set the parameters and judge the output.
Concentrix CEO Chris Caldwell put it bluntly: you don't need a hundred Jordans running around. You need people who can manage AI like it's an entire team.
That's orchestration. And it requires context engineering. The AI needs everything upfront before it goes deep. Otherwise you're just hoping it guesses right.
Try This:
Open your most-used AI tool and create a dedicated project for your single most repeated task.
Dump everything the AI needs: your role, your standards, your edge cases, your preferred format.
You'll cut the back-and-forth in half because the AI finally has context instead of you explaining yourself from scratch every time.
3. Unlearn or fall behind 🔄
This is where it gets uncomfortable.
"Upskilling" has set companies back years. It treats AI like a course correction. Just add new tricks to old habits.
That's backwards.
Organizations seeing ROI triple aren't upskilling. They're unlearning. Throwing away processes that worked for decades and rebuilding from scratch with AI at the center.
You can't throw makeup on an ugly process and expect it to be pretty. It's still ugly. Now it just has makeup.
The teams winning killed their old playbooks. You should too.
Try This:
Pick one workflow your team has done the same way for years.
Sketch that version out this afternoon.
The rebuild probably looks nothing like your current process. That's the point.
Reply