• Everyday AI
  • Posts
  • How to put users first in building AI 🥇

How to put users first in building AI 🥇

🧐 Impact of AI on product strategy, Google Gemini video was kinda faked, Web Reader ChatGPT plugin review, and more!

Author

Everyday AI
December 08, 2023

Outsmart The Future

Today in Everyday AI
7 minute read

🎙 Daily Podcast Episode: How can we create better user centric AI to help with product strategy? What effect will it have on us as consumers? Give it a listen.

🕵️‍♂️ Fresh Finds: An interview AI copilot, Toyota using AI for its HR and IT, and senior citizens using AI as companions. Read on for Fresh Finds.

🗞 Byte Sized Daily AI News: Google Gemini demo video wasn’t super truthful, Meta’s new Purple Llama, and EU act talks continue. For that and more, read on for Byte Sized News.

🚀 AI In 5: We’re highlighting a ChatGPT plugin that is a MUST have in your workflow! We’re showing you why. See it here.

🧠 Learn & Leveraging AI: So what goes into user centric AI and where can you start? Keep reading for that!

↩️ Don’t miss out: Did you miss our last newsletter? We talked about Using AI and podcasting for growth, EU close to AI agreement, and Doc Maker ChatGPT plugin. Check it here!

Product Strategy in the Age of AI 🤔

We’re constantly being thrown products left and right. And with the AI wave, AI centered products are also everywhere we look.

How can we create better AI that's centered around users?

What influence will AI have on products and its users?

Svetlana Makarova, AI Group Product Manager at Mayo Clinic, joins us to discuss how AI will reshape product strategy and management.

Also on the pod today:

• Importance of explainable AI 🇦🇺
• Creating user centric AI 🗣
• Implementing AI in product strategy 🤖

It’ll be worth your 32 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 – Final Round is an interview copilot, Birdie is an all-in-one social media growth tool, and Outboundly.ai is an AI-powered cold outreach platform.

Money in AI – A US banking regulator is warning financial firms about the risks AI poses.

Business of AI - Toyota Motor North America is using AI for its HR and IT service desk requests.

Read This – Stanford University researchers are saying that tech firms are failing to use ethical AI.

AI in Society – Senior citizens in New York State are using AI to combat loneliness.

1. Google's Gemini AI Video wasn’t super truthful 🤥

Google's recent launch of Gemini, its powerful AI model suite, has been met with skepticism over its demo video's authenticity. Critics, including a Bloomberg columnist, argue that the video misrepresented Gemini's real-time capabilities, as Google admitted to reducing latency and shortening outputs for the demo.

The video showcased Gemini's multimodal abilities, like image recognition and conversational prompts, but was later revealed to use still image frames and written text prompts, not real-time voice interaction.

2. AMD's AI Chip Market Forecast 📈

AMD has significantly raised its market forecast for AI chips to a staggering $400 billion by 2027. This optimistic projection suggests a rapidly expanding market for AI hardware, with AMD poised to increase its market share. However, Nvidia, the current industry leader in AI chips, stands to benefit from this growth as well. Nvidia's established position and technological advancements position it well to handle the increasing competition in the AI chip market.

3. Meta's 'Purple Llama': Enhancing AI Safety and Responsibility 🛡️

Meta has introduced "Purple Llama," a project linked to its Llama AI tool, focusing on safety tools and evaluations for responsible AI development. This initiative combines attack (red team) and defensive (blue team) strategies, termed "purple teaming," to evaluate and mitigate AI risks. Purple Llama, which includes the widely downloaded Llama 2 model, offers permissive licensing for both research and commercial use, fostering collaboration and standardizing trust and safety tools.

4. EU's AI Act Talks: Balancing Innovation and Security 🇪🇺

The European Union's negotiations on AI regulations enter a third day, focusing on military and security applications. After a marathon debate, lawmakers reached a provisional deal on regulating AI systems like ChatGPT. The main discussion now centers on the use of AI in biometric surveillance, with governments advocating for exceptions in national security, while lawmakers push for a ban due to privacy concerns.

5. Amazon's AI Vision: For Customer Experiences 🛍️

Amazon CEO Andy Jassy predicts that generative AI will revolutionize customer experiences across the company's enterprise and consumer sectors. In an interview with CNBC, Jassy emphasized the potential of AI to enhance Amazon's predictive capabilities and improve Alexa's functionality. He acknowledged the risks associated with AI, advocating for investment in security and caution against AI-generated inaccuracies.

WebReader ChatGPT plugin review

Welcome back to day 2 of the 12 days of plugins!

Today's plugin is Web Reader.

This is one of those ChatGPT plugins that is a MUST for your workflow! We're showing you why!

🤷‍♂️ What’s Going On and Why It Matters:

If you’re working in product management or strategy, the go-to might just be ‘sprinkle some AI on that.’ 

Not so fast. 

And even if you don’t work in product, this convo’s an important one to take in. 

Svetlana Makarova joined the Everyday AI show to talk all things product strategy and how focusing on the users is more important than focusing on the product. 

And Svetlana knows a thing or two. She serves as the AI Group Product Manager at the Mayo Clinic, known for its innovation in patient care and leader in the intersection of AI and healthcare. 

(See their partnership with Google Cloud for GenAI and integration with Microsoft for GenAI tools.) 

In other words, Svetlana was the perfect person to talk about approaching AI product development the right way: user-centric. 

Here’s a sampling of what we covered: 

  • Decision-Making Process for Implementing AI

  • Importance of measuring ROI before scaling use cases

  • Evolution of user-centric AI and understanding user workflows for task automation

  • Focusing on the biggest return on investment

  • Automating tasks through understanding data and user feedback

  • Core components for scaling AI: data, platform approach, flexible infrastructure

And you guessed it — there’s a lot more. 

And today’s convo might impact you more than you know. 

Let’s chat about it 🗣️

🦾How You Can Leverage:

Product strategy impacts us all.

The computer or phone you’re reading this on. The browser you may be in right now. That loud noise coming from the other room (Hey.. what the?)

In short, products impact every aspect of our daily life. And with the nonstop rush to integrate AI into just about every product, offering and service, it’s important to understand a user-centric approach to it all.

1 – AI should be invisible️ 👻

Wanna make sure your product or service is seamless? 

Svetlana says AI implementation should be invisible for the user. Like… oh! Look at that AI. 

In the age of AI, the goal is for AI to seamlessly blend into user experiences, almost becoming invisible.

Try this: 

Just like Svetlana mentioned about Google's integration of AI into search, the focus should be on enhancing results without disrupting the user's journey. 

That leads to improved user satisfaction and engagement. Win-win. Wanna dive even deeper? Check out this article on seamless AI integration from akeneo. 

2 – User-centric = explainability 🔣

If you can’t explain why you’re using AI or how it works to the end user, then you shouldn’t use it. 

Simple as that shorties. 

Svetlana said that prioritizing the explainability of AI-driven processes and decisions, businesses can build trust with users, foster transparency, and mitigate potential concerns related to AI technology.

Try this: We tackled the explainability piece in more depth a few weeks ago with a former McKinsey leader. Go check that out! 

3 – Three steps for scaling AI 3️⃣

Svetlana shared her own 3 steps for scaling products with AI? 

Sheeeeesh!

She must know our affinity for all simple lists of three. Svetlana mentioned the 3 core components for scaling AI: data, platform approach, flexible infrastructure. 

Try this: 

It’s like a bonus round for #3, straight from the source: 

Step 1: Svetlana emphasizes the importance of having access to high-quality, diverse datasets to train AI models effectively. Without good data, the AI can't learn and improve, rendering the whole endeavor pointless.

Step 2: Platform approach - Svetlana underlined the need for a comprehensive platform for developing AI applications. This platform should facilitate the entire AI development process, from data collection and model training to deployment and ongoing monitoring.

Step 3: Flexible infrastructure - Svetlana stresses the significance of flexible infrastructure for experimentation. This allows for testing and refining AI solutions without disrupting existing operations, making the scaling process smoother and more efficient.

Now This …

What do you think?

Vote to see live results

Do you feel overwhelmed by AI products?

Login or Subscribe to participate in polls.

Reply

or to participate.