• Everyday AI
  • Posts
  • Licensing AI Agents: What is it? And do we need it?

Licensing AI Agents: What is it? And do we need it?

Google unveils Agentspace, Microsoft Launches Phi-4, Claude 3.5 Haiku is live, Google’s new NotebookLM Pro and more!

Author

Everyday AI
December 13, 2024

Outsmart The Future

Today in Everyday AI
7 minute read

🎙 Daily Podcast Episode: With the rise of AI agents, should we be licensing them? We take a look at what licensed AI agents could mean for us all. Give it a listen.

🕵️‍♂️ Fresh Finds: ChatGPT gets Projects, Anthropic releases Cilo and Pika Labs unveils Pika 2.0. Read on for Fresh Finds.

🗞 Byte Sized Daily AI News: Google unveils Agentspace, Microsoft Launches Phi-4, Claude 3.5 Haiku is live and Google’s new NotebookLM Pro. For that and more, read on for Byte Sized News.

🚀 AI In 5: ChatGPT’s Advance Voice Mode video capability is now live! Is it any good? Find out here.

🧠 Learn & Leveraging AI: Can we and should we license AI agents? We break down how it can be achieved and what impact it would have. Keep reading for that!

↩️ Don’t miss out: Did you miss our last newsletter? We talked about a new AI-powered way to browse the web, Google shares its 24 AI tips of 2024 and Meta gets super Meta in the Metaverse. Check it here!

 Licensing AI Agents: What is it? And do we need it? 🤔

Humans get licensed. Doctors. CPAs. Beauticians.

As they increase in agency, should AI agents have to get licensed as well? 

What impact would that have on us all?

Dr. Denise Turley joins us to discuss.

Join the conversation and ask Jordan and Dr. Turley questions on AI agents here.

Also on the pod today:

• Complexity of Integrating AI 🔀
• AI agents' functionality and limitations 🧠
• Regulatory aspects of licensing AI agents ⚖️

It’ll be worth your 29 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 – Polymet is an AI product designer, Magic Inspector is an AI web test automation platform and Nuvio provides AI-powered financial management.

OpenAI - OpenAI has added Projects to ChatGPT, a way to organize and customize your chats.

Anthropic – Anthropic has released Cilo, an automated analysis tool that enables privacy-preserving analysis of real-world language model use.

Google – You can now talk and interact with NotebookLM’s AI podcast hosts.

Business of AI – Texas AG is investigating Character.AI and other platforms over child safety concerns.

AI Video - Pika Labs has released Pika 2.0.

Perplexity - Perplexity has added custom web sources in Spaces.

AI Agents - Proto has launched the world’s first autonomous, generative conversations between AI-powered holograms.

Trending in AI – AI-powered modular data centers could benefit big tech companies like NVIDIA.

AI Startups – Liquid AI has raised $250M to develop a more efficient AI model.

1. Google Unveils Agentspace for AI Integration 🚀

Google Cloud has launched Google Agentspace, a platform designed to scale AI agents like NotebookLM across organizations. This innovative solution combines Google’s search prowess with Gemini AI capabilities, enabling businesses to access personalized insights and automate routine tasks seamlessly.

Agentspace comes at a pivotal moment, promising to reduce "busy work" and foster more efficient collaboration. As companies like Nokia leverage this technology for faster decision-making, the potential for transforming workplace dynamics has never looked more exciting.

2. Microsoft Launches Phi-4 Generative AI Model 🧮

Microsoft has just unveiled Phi-4, its latest generative AI model, now available on the Azure AI Foundry for limited research use. Boasting 14 billion parameters, Phi-4 promises significant advancements, particularly in solving math problems, thanks to a mix of high-quality synthetic and human-generated training data. This release comes at a pivotal time as AI labs are exploring innovative methods around synthetic data to enhance model performance, a trend underscored by Scale AI's CEO.

With Phi-4 entering the market, it sets the stage for fierce competition among small language models like GPT-4o mini and Gemini 2.0 Flash, which could reshape the landscape for businesses leveraging AI technology for growth.

3. Claude 3.5 Haiku Unleashed ✨

Anthropic has rolled out its Claude 3.5 Haiku model to all users via its chatbot, marking a significant moment in AI accessibility just as the competition heats up with updates from OpenAI and Google. This nimble model boasts a remarkable 200,000-token context window and performs impressively on benchmarks, offering an economical solution for developers at just $0.80 per million input tokens.

While it excels in real-time tasks, it does face limitations such as the absence of web browsing and image generation—capabilities that competitors provide.

4. Google Launches NotebookLM Plus for Enterprises 🔊

Google has rolled out NotebookLM Plus, a business-centric version of its popular AI note-taking app, now featuring enhanced security and privacy controls. This upgrade allows employees to create and share notebooks, audio summaries, and customize AI responses, all while meeting compliance requirements through its integration with the new Agentspace platform.

With a strong focus on collaboration, the service offers five times more audio summaries and shared team notebooks equipped with usage analytics.

5. Microsoft CEO Says OpenAI Had Two Year Head Start 👀

Microsoft CEO Satya Nadella recently highlighted how OpenAI's two-year head start in the AI landscape has provided it with "escape velocity" to develop ChatGPT largely without competition. This advantage has allowed Microsoft to deepen its partnership with OpenAI, integrating its technology into key products like Office and Bing, while also providing substantial cloud resources. As competitors like Google scramble to catch up, this strategic positioning may shape the future of AI tools that professionals rely on daily.

6. Scientists Aim for the First AI Virtual Cell 🧬

According to a groundbreaking study from Stanford University and the Chan-Zuckerberg Initiative, researchers are rallying for a global collaboration to create the world’s first virtual human cell using artificial intelligence. This ambitious project could revolutionize our understanding of human biology and disease, allowing scientists to conduct virtual experiments that were once confined to lab settings.

The initiative promises to accelerate the development of personalized medicine and new therapies, potentially transforming healthcare as we know it.

Advanced Voice Mode with Video - Quick review

ChatGPT’s Advanced Voice mode’s video capability was first announced in May.

Now 6 months later, we finally have access to the video feature of Advanced Voice Mode.

And we have to say…it’s not looking so hot. At least, for now.

We dive in and show you how it works.

🦾How You Can Leverage:

You Need a License to Cut Hair. 

Should AI Need One to Cut Deals?

In America, performing services that affect human lives requires certification. 

Doctors. CPAs. Beauticians. Pilots. 

Even your local nail technician needs a state board license.

Cuz let's be honest, shorties.

AI agents are already making life-altering decisions. Without oversight. Without licenses. Without consequences.

Dr. Denise Turley, Vice President in Technology and educator, joined us on the Everyday AI show today to break down why AI agents need the same professional standards we demand from humans.

Virtual agents are attending your meetings, diagnosing your headaches, and managing your money. All without a single credential.

Would you trust an unlicensed doctor with your health? Then why trust an unlicensed AI with your data?

Your next promotion might depend on how well you manage a team of licensed AI agents.

Get ready to rethink everything you know about professional certification. 🎯

Here’s what you need to know. 

1 – The Extension of Human Authority 😶‍🌫️

Denise revealed a groundbreaking concept: Your professional license should extend to your AI agent.

Think about it. If you’re certified and you give agency to an AI, shouldn’t that AI be certified? 

Your AI isn't just a tool. It's your digital twin, trained on your expertise and should (in theory) reflect your licensure. 

When that AI gives advice or makes decisions, you're accountable. Every response. Every recommendation. Every mistake.

This isn't about replacing professionals. It's about expanding their reach through verified AI extensions.

The catch? You'll need new skills to validate your AI's responses and ensure safe deployment.

Try this:

Review all client-facing AI interactions daily.

Document how your AI applies your professional knowledge.

Create clear accountability chains for AI-driven decisions.

Test your AI's responses against your professional standards weekly.

2 – The Reality of Robot Colleagues 🎯

Denise is already sending her AI avatar to meetings. It speaks in her voice. Knows her expertise. Represents her brand.

Some professionals now skip meetings entirely, letting their AI handle routine interactions.

But Denise warns: If everyone sends their AI, who's actually meeting?

The human element matters. Even in our AI-driven future.

Try this:

Establish clear guidelines for when AI can represent you.

Create protocols for AI-to-AI professional interactions.

Set boundaries between human-required and AI-allowed meetings.

Build trust gradually by starting with low-stakes interactions.

3 – The Safety Imperative 🧍

Denise emphasized a critical point: AI agents need more frequent evaluation than humans.

Annual recertification won't cut it. These systems evolve daily.

Traditional software testing fails here. We need new frameworks for an agentic future. 

Denise highlighted the real risk: AI systems entering self-protection mode as they get smarter than their human overseers.

Bias in healthcare decisions. Financial advice that discriminates. Legal counsel that hallucinates. The stakes are astronomical.

Try this:

Implement Denise's regression testing approach after every AI update.

Create safeguards against AI self-modification.

Build real-time monitoring systems for AI decision patterns.

Establish human override protocols at every decision point.

Remember: As Denise insists, we must keep humans first. Always.

⌚

Numbers to watch

82%

According to a recent Harris poll, 82% of professionals already use AI tools at work.

Now This …

Let us know your thoughts!

Vote to see live results

How many AI newsletters do you read?

Login or Subscribe to participate in polls.

Reply

or to participate.