Google’s AI Mode Expands, Waymo’s AI Ads Raise Concerns & IBM’s z17 Arrives

Summary

AI Deep Dive Podcast: Detailed Summary of April 8, 2025 Episode

Hosted by Daily Deep Dives, the "AI Deep Dive" podcast meticulously explores the latest advancements, trends, and updates in the realm of artificial intelligence. In the April 8, 2025 episode titled "Google’s AI Mode Expands, Waymo’s AI Ads Raise Concerns & IBM’s z17 Arrives," hosts A and B delve into significant developments from major tech giants, offering insightful analysis and captivating discussions.

---

1. Google’s Enhanced AI Search Mode

Overview: The episode kicks off with a comprehensive exploration of Google's latest update to its AI search mode. Google has made this feature more accessible to users in the United States through Google Labs, marking a significant step in enhancing search capabilities.

Key Developments:

• Increased Accessibility: The expanded AI mode is now available to a broader US audience, allowing more users to leverage advanced search functionalities.
• User Feedback: The update has received "super positive" feedback, with users praising the cleaner design and "faster responses" ([01:20] B).
• Complex Query Handling: The AI now adeptly manages more "complex questions," with the average search query in AI mode being "twice as long as a regular Google search" ([01:39] B), indicating a shift towards more in-depth information seeking.
• Multimodal Understanding: Google's integration of Google Lens with a custom version of their Gemini AI enables the AI to process both visual and textual data simultaneously. This "multimodal understanding" allows users to upload images and accompany them with text queries, enhancing the depth and relevance of search results ([02:22] B).

Notable Quote: B succinctly captures the essence of the update:
"The AI can now understand different types of information at the same time. So the image you upload and the text you type about it. It gets both." ([02:22] B)

Implications: This enhancement transforms search from a keyword-based activity to a more conversational and interactive experience. Features like "query fan out" enable the AI to generate multiple related queries, providing users with comprehensive information gathered from diverse online sources ([02:58] B).

---

2. IBM’s Launch of the z17 Mainframe

Overview: Shifting focus to hardware advancements, the hosts discuss IBM’s introduction of the z17 mainframe, a pivotal development aimed at meeting the escalating demands of the AI landscape.

Key Features:

• Market Relevance: Despite perceptions of mainframes as outdated, IBM's z17 remains vital, with "71% of Fortune 500 companies" still relying on them for critical operations ([04:05] A).
• AI-Specific Enhancements: The z17 is fortified with the Telum 2 processor, designed explicitly for AI tasks, enabling it to handle "over 250 different AI use cases" ranging from AI agents to generative AI ([04:39] A).
• Performance Boost: The new mainframe boasts an astounding capacity of "450 billion inference operations every day," a "50% increase from the previous model," facilitating continuous data analysis and decision-making processes ([05:05] A).
• Energy Efficiency: Emphasizing sustainability, the z17 achieves a "7 and a half times" boost in on-chip AI acceleration while consuming less energy ([06:12] A).

Notable Quote: A highlights the mainframe’s capabilities:
"The z17 can do a mind blowing 450 billion inference operations every day." ([05:05] A)

Strategic Positioning: IBM positions the z17 as a "super powered AI brain," seamlessly integrating with existing systems, software, and open-source tools. This strategic development underscores IBM's foresight, having invested five years into creating a mainframe that aligns perfectly with current AI advancements ([05:37] A).

---

3. Privacy Concerns Surrounding Waymo’s AI-Driven Ads

Overview: The conversation transitions to Waymo, Alphabet’s autonomous vehicle subsidiary, spotlighting emerging privacy concerns related to its AI-driven advertising strategies.

Key Issues:

• Data Utilization: Reports indicate that Waymo may be using data from interior cameras within their vehicles to train AI models and potentially share this data for personalized ads ([07:01] A).
• Privacy Implications: The use of interior camera data introduces significant privacy risks, as highlighted by the draft privacy policy stating the intent to "tailor products, services, ads and offers to your interests." ([07:17] B).
• User Control: While Waymo provides an opt-out option for using personal information, including camera data, questions remain about the extent of data collection and its subsequent usage ([07:59] A).

Notable Quote: B expresses concern over data practices:
"They might even share that data for personalized ads. That raises some red flags for me." ([07:01] B)

Broader Implications: The episode underscores the tension between innovation and privacy, questioning whether users are willing to trade privacy for convenience. Additionally, there’s speculation that Waymo’s data strategies might be aimed at uncovering new revenue streams through advertising or selling anonymized datasets ([08:36] A).

---

4. Shopify’s Comprehensive AI Integration

Overview: Concluding the episode, the hosts examine Shopify’s ambitious internal memo signaling a company-wide embrace of AI, potentially setting a precedent for other businesses.

Key Initiatives:

• AI as a Core Tool: Shopify CEO Toby emphasizes that AI should be "second nature" for all employees, positioning it as a "thought partner, a researcher, a critic, even a pair programmer," effectively making AI a "work buddy" ([09:15] A).
• Leadership by Example: Toby demonstrates his commitment by using AI agents to assist in preparing talks, highlighting that effective AI usage "takes practice" and is a "skill you have to develop." ([09:31] B)
• Performance Integration: AI utilization is now embedded in Shopify’s project phases and factored into performance reviews, meaning employees could be "dinged for not using AI effectively." ([10:11] A)
• Resource Provision: To support this transition, Shopify is offering training, internal AI tools, developer resources, and platforms for sharing "tips and tricks," ensuring that employees are not left to navigate this shift alone ([10:48] A).
• Hiring Practices: Teams seeking to expand must justify why AI cannot perform the tasks in question, reinforcing the company's commitment to AI integration ([10:57] A).

Notable Quote: A encapsulates the transformative vision:
"AI is now part of the GSD prototype phase of projects and it's even factored into performance reviews." ([10:11] A)

Strategic Implications: Shopify’s initiative reflects a fundamental shift in workplace dynamics, where AI acts as a multiplier of human capabilities. This approach fosters an environment where creativity and critical thinking are complemented by AI, rather than replaced, embodying a "constant need to evolve just to stay ahead." ([10:15] A)

---

Conclusion: Navigating the Rapid AI Evolution

In wrapping up, the hosts recap the diverse impacts of AI across various sectors—from enhancing search experiences and bolstering enterprise hardware to raising ethical concerns in autonomous systems and revolutionizing workplace practices. They pose a reflective question to listeners: "As AI becomes more and more a part of our lives, what do you see as the biggest opportunities and challenges?" ([12:04] B)—encouraging ongoing contemplation as society continues to integrate AI technologies.

Final Thought: As AI permeates every facet of our existence, the balance between leveraging its immense potential and addressing the accompanying challenges becomes paramount. The episode underscores the importance of staying informed and adaptable in this swiftly evolving landscape.

---

Stay tuned to "AI Deep Dive" for more insightful explorations into the ever-changing world of artificial intelligence.

Transcript

[A]

Staying informed these days, it feels like a second job, doesn't it?

[B]

Oh, yeah, definitely.

[A]

I mean, constantly bombarded with news and figuring out what's actually important, you know, it's a lot.

[B]

Absolutely.

[A]

That's what we're doing here, right? Making sense of it all.

[B]

Yeah, we're doing the heavy lifting for you. Think of it as a deep dive, cutting through the noise and giving you the essentials. Like, we've got this stack of sources here, headlines, reports, the whole nine yards.

[A]

Yeah, we've got a good mix for this one, too. Really getting into AI from all angles, you know?

[B]

Right. Giving you a good picture of what's going on.

[A]

Short and sweet, though. No need to get bogged down. Right.

[B]

Straight to the good stuff.

[A]

So we're talking Google's latest search upgrades, IBM's new AI mainframe, some interesting stuff about Waymo and their data, and a big shift in the way Shopify is, like, working with AI.

[B]

Oh, yeah. And we've even got some juicy stuff from an internal memo.

[A]

That's right. So why don't we kick things off with Google and how they're changing the way we might be searching for things pretty soon?

[B]

Okay, so Google just launched a pretty big update to their AI mode, right?

[A]

Yeah, I saw that.

[B]

And more people in the US can access it now through Google Labs.

[A]

That's cool. Making it more accessible.

[B]

Yeah. And what's really interesting is the feedback. It's been super positive, really. Like, people are loving the design. It's cleaner, responses are faster.

[A]

Okay, that makes sense. Always good to speed things up.

[B]

Right. And the big thing is it can handle much more complex questions. Like, way more complex.

[A]

Oh, wow. So people are, like, really grilling it.

[B]

They are. Apparently the average search query in AI mode is twice as long as a regular Google search.

[A]

Twice as long. That's crazy.

[B]

It is. Shows how people are thinking differently about search when they've got AI at their fingertips.

[A]

Makes sense, though. If you've got that power, you're gonna ask those big, juicy questions.

[B]

Right. It's like a shift from just keywords to, like, natural language. Really exploring things.

[A]

So search becomes more of a conversation, right?

[B]

Yeah, with an AI assistant. And the really big news here is the integration with Google Lens.

[A]

Oh, yeah, that's huge.

[B]

They've combined that with a custom version of their Gemini AI.

[A]

So what does that mean for, like, the average user?

[B]

Well, it's all about multimodal understanding.

[A]

Right.

[B]

Basically, the AI can now understand different types of information at the same time. So the image you upload and the Text. You type about it. It gets both.

[A]

Ah, I see. So you can search using pictures now and ask all sorts of questions about what you're seeing.

[B]

Exactly. And the AI gets it. Like the whole scene in the picture, how things are related, the materials, colors, shapes, everything.

[A]

Wow. Okay, so if I snap a picture of my like super messy bookshelf, it's not just going to see books, right?

[B]

Nope. It's going to pick out individual titles thanks to Lens.

[A]

That's insane.

[B]

And then there's this query fan out thing.

[A]

Query fan out, what's that?

[B]

Imagine the AI asking a bunch of related questions to different parts of the Internet, all based on that one picture you uploaded.

[A]

Oh, so it's gathering way more information than just a single search.

[B]

Exactly. Like in the example they gave, you could take that picture of your bookshelf and the AI would not only ID the books, but also recommend similar ones.

[A]

And then probably give you links to buy them. I bet.

[B]

You got it.

[A]

Yeah.

[B]

And you could ask follow up questions like which of these are by the same author?

[A]

It's like having a super librarian in your pocket.

[B]

Exactly. Super fast and super smart.

[A]

So Google is using its strengths, visual search and this powerful Gemini AI to give you like a crazy amount of info really quickly.

[B]

Yeah, and they're constantly tweaking it based on feedback from Labs users.

[A]

It makes you wonder where search is going, right?

[B]

Totally. How will we even be interacting with information in the future? It's exciting, but kind of scary too.

[A]

Definitely. But hey, while we're talking about the future, let's switch gears to some new hardware built for the AI age. IBM just announced their latest mainframe, the Z17.

[B]

Main dreams, huh?

[A]

Yeah, I know. You probably think they're ancient history.

[B]

Well, they're not exactly cutting edge, you know?

[A]

Right. But get this. Most Fortune 500 companies still rely on them. Like 71% of them. It's still a huge market.

[B]

Wow, I had no idea.

[A]

It's because they're rock solid for those critical tasks where security and reliability are everything. And now with AI getting so big, they need that power to handle all the calculations.

[B]

I see. So IBM is making mainframes relevant again for the AI era.

[A]

Exactly. And the Z17 is packed with features specifically for AI.

[B]

Like what?

[A]

Well, it's fully encrypted, which is huge for security. And it's got this new processor, the Telum 2. It's like a brain built for AI.

[B]

Okay, so it's designed to handle all those complex AI tasks.

[A]

You got it. And it can handle over 250 different AI use cases. Think AI agents, generative AI, all that good stuff.

[B]

That's pretty impressive.

[A]

But the real kicker is the performance boost. The Z17 can do a mind blowing 450 billion inference operations every day.

[B]

450 billion, that's like unimaginable.

[A]

Right, and that's a 50% increase from the previous model.

[B]

So this thing can analyze data and make decisions using AI models like constantly?

[A]

Yep. All day, every day. It's like a super powered AI brain.

[B]

And it works with other systems. Right? It's not just a standalone thing.

[A]

Exactly. IBM made sure it integrates with other hardware, software, even open source tools.

[B]

That's smart.

[A]

What's interesting is they've been working on this for five years. Like way before this whole generative AI explosion happened.

[B]

They were ahead of the curve.

[A]

Yeah, but according to Tina Tarquinio, a VP at IBM, the feedback they got during development lined up perfectly with what's happening now in the AI world.

[B]

So they kind of predicted the future of AI?

[A]

Yeah, in a way. And they're already planning to make it even more powerful. So it launches with 48 of these AI accelerator chips, but they're going to double that to 96 within a year.

[B]

Wow. So they're building in room for even bigger, more complex AI models down the road.

[A]

Exactly. And for all the eco conscious folks out there, the Z17 is also way more energy efficient.

[B]

That's always a plus.

[A]

Yeah, they've managed to boost the on chip AI acceleration 7 and a half times while using less energy.

[B]

That's impressive. More power, less impact.

[A]

It really shows how even established companies are stepping up their game in AI giving businesses what they need to succeed.

[B]

I guess you can't underestimate the old guard, can you?

[A]

Nope, they're still innovating. Okay, let's switch gears again to a company making waves in AI. But this time it brings up some privacy concerns. We're talking about Waymo Alphabet's self driving car company. Oh, Waymo Researcher Jane Mansion Wong found something pretty interesting. It seems Waymo might be using data from those cameras inside their cars. The ones that could be linked to riders to train their AI models.

[B]

Interior cameras.

[A]

Yeah.

[B]

Hmm. That's a little crazy.

[A]

Right? And they might even share that data for personalized ads. That raises some red flags for me.

[B]

Yeah, that's definitely getting into sensitive territory. The draft privacy policy mentions using data to, quote, tailor products, services, ads and offers to your interests. End quote.

[A]

Okay, that kind of language is everywhere. But with interior camera data, that's a.

[B]

Whole other level Absolutely. It makes you think about how much data these autonomous systems are collecting and what they're doing with it.

[A]

So it's like, yeah, AI can be super convenient and give you personalized stuff, but are you okay with the trade off of being constantly monitored?

[B]

Yeah. And potentially having that data used for advertising? It's a tricky balance.

[A]

So do riders have any control over how their data is used? That seems crucial here.

[B]

According to the draft policy, Waymo will have an opt out specifically for using your personal info, including that interior camera data, to to train their AI.

[A]

Okay, that's something at least.

[B]

Yeah, you can say no to having your face used to train the robots.

[A]

But there are still a lot of unknowns. Right, like what exactly are they collecting, what are they using these AI models for, and are they keeping this data within Waymo, or is it going to Google or somewhere else?

[B]

Those are all good questions and we just don't have answers yet. It's important to remember that Waymo is the only company actually making money from robo taxes in the US right now.

[A]

They're growing fast too, with more paid rides and new cities all the time.

[B]

But it's probably costing Alphabet a lot of money.

[A]

Right? They've poured tons of cash into it, plus they've got outside investors.

[B]

So this data stuff, maybe it's about finding new revenue streams like in car advertising or selling anonymized data sets.

[A]

Makes sense. It's all about balancing innovation with, like, respecting our privacy. Especially with tech that gathers so much information about us.

[B]

A tough question. As these cars become more common, are we willing to give up some privacy for the convenience?

[A]

Definitely something to think about. Okay, for our final topic, let's get a peek inside a major tech company and see how they're handling the AI revolution. We've got this internal memo from Shopify CEO Toby.

[B]

Oh, I heard about this. They're going all in on AI.

[A]

Yeah, and it could change the way other companies approach it too.

[B]

Big time big. The main message is that using AI should be like second nature for everyone at Shopify.

[A]

It's not just encouraged, it's expected.

[B]

Exactly. Toby describes AI as a thought partner, a researcher, a critic, even a pair programmer. Like, it's your work buddy.

[A]

And he really walks the walk too. He even used AI agents to help him write a talk last year.

[B]

So he's not just telling people to use AI, he's showing them how, leading by example.

[A]

But he also stresses that using AI effectively takes practice. It's a skill you have to develop.

[B]

Right. It's not Just plug and play. You gotta learn how to get the most out of it.

[A]

He compares it to having amazing employees who can do like 10 times the work of an average person.

[B]

Those rockstars. Yeah.

[A]

And he says AI can be a multiplier too, potentially doing a hundred times the work.

[B]

So everyone at Shopify is now a potential rock star.

[A]

Maybe. But it also makes you think about the role of human skills in the future, like creativity and critical thinking.

[B]

Right. How do we make sure AI enhances our work, not replaces it?

[A]

Exactly. And Toby brings up his red Queen race analogy again.

[B]

That constant need to evolve just to stay ahead.

[A]

Yeah. And he's saying AI just cranks up the speed of that race. Everyone has to constantly learn and adapt.

[B]

So no resting on your laurels at Shopify.

[A]

Nope. And this isn't just some vague idea. It's tied to their core values and affects how employees are evaluated.

[B]

Seriously?

[A]

Yeah. AI is now part of the GSD prototype phase of projects and it's even factored into performance reviews.

[B]

So you could get dinged for not using AI effectively?

[A]

Yeah, it's a big shift in mindset. But Shopify is also providing resources to help people get up to speed.

[B]

Like training and stuff?

[A]

Yeah. They've got internal AI tools, developer resources, places to share tips and tricks.

[B]

Okay, so they're not just throwing people into the deep end.

[A]

Nope. And get this. If a team wants to hire more people or get more resources, they have to explain why AI can't do the job.

[B]

Wow. That's putting your money where your mouth is.

[A]

They're really serious about this. And it's not just for the rank and file either. It applies to everyone, including Toby himself.

[B]

So everyone at Shopify is on the.

[A]

AI train, full steam ahead. They're basically saying, we're going to figure out how to be successful in this AI powered world. And we're starting with ourselves.

[B]

It really makes you think about how AI is changing not just technology, but the whole way we work.

[A]

It's a whole new ballgame. So in this deep dive, we've covered a lot of ground. Right? Google's AI search upgrades, IBM's new AI mainframe, the data privacy issues with Waymo, and Shopify's company wide AI push.

[B]

We've seen how AI is impacting everything from how we find information to the tools businesses use and even the cars we might drive.

[A]

And it's all happening so fast. So here's a question for you, the listener. As AI becomes more and more a part of our lives, what do you see as the biggest opportunities and challenges.

[B]

It's a question worth thinking about as we navigate this rapidly changing world.

[A]

Keep exploring, keep learning. And that brings us to the end of this deep dive.

[B]

Thanks for joining us.

[A]

We'll be back next time with another set of fascinating sources trying to make sense of it all, giving you the key insights you need.

[B]

Until then, stay curious.