Google’s Gemini 2.0 Pro & ‘Ask for Me’, OpenAI’s Lab Partnership, and Apple’s AI Strategy

Summary

AI Deep Dive: Episode Summary Hosted by Daily Deep Dives

Episode Title: Google’s Gemini 2.0 Pro & ‘Ask for Me’, OpenAI’s Lab Partnership, and Apple’s AI Strategy
Release Date: January 31, 2025

---

1. Introduction

In this episode of the AI Deep Dive podcast, hosts A and B navigate through the latest and most impactful developments in the artificial intelligence landscape. The conversation delves into Google's new AI model, OpenAI's strategic collaborations, and Apple's evolving AI strategies, while also addressing the ethical and societal implications of these advancements.

---

2. Google’s Latest AI Developments

a. Gemini 2.0 Pro Experimental

Google has recently unveiled its flagship AI model, Gemini 2.0 Pro Experimental. Despite Google's prominence in the AI sector, the release was notably understated, with updates primarily found within the Gemini app's changelog.

Speaker A [00:26]: "You wouldn't even know it unless you really dug into the changelog for their Gemini app."

The new Gemini upgrade emphasizes enhanced factual accuracy and improved performance in coding and mathematical tasks. However, its accessibility is currently limited to users subscribed to Google's premium Gemini Advanced service, indicating a cautious and phased rollout approach.

Speaker B [01:04]: "It offers better factuality... stronger performance for coding and math related tasks."

This selective release strategy suggests Google's intent to refine the model further before a broader public deployment, mitigating the risk of potential AI inaccuracies.

b. 'Ask for Me' Feature

In addition to Gemini 2.0 Pro, Google is introducing the 'Ask for Me' feature, an AI-driven assistant designed to handle everyday tasks on behalf of users. This feature can, for instance, schedule appointments, inquire about services, and provide summaries of interactions, significantly streamlining routine activities.

Speaker B [01:58]: "It contacts businesses on your behalf. It finds out about their services, their pricing, their availability and all that."

While this innovation promises substantial time savings, it also raises concerns about the displacement of jobs that involve similar tasks, such as customer service representatives.

Speaker A [02:26]: "But you know, it does make you think what happens to the people whose jobs involve making those calls."

The hosts acknowledge the need to address the workforce impact as AI continues to integrate into more facets of daily life.

---

3. OpenAI’s Strategic Partnerships

OpenAI has entered into a significant collaboration with the US National Laboratories, deploying its reasoning models on the Venato supercomputer at Los Alamos National Laboratory. This partnership aims to harness AI for tackling monumental challenges across various sectors.

Speaker B [02:45]: "They're aiming to tackle some really big challenges... disease treatment and prevention, cybersecurity, energy, innovation."

Key focus areas include:

• Healthcare: Enhancing disease treatment and prevention strategies.
• Cybersecurity: Strengthening defenses against cyber threats.
• Energy: Innovating sustainable energy solutions.
• Nuclear Security: Reducing risks associated with nuclear materials and technology.
• Astrophysics: Deepening our understanding of the universe.

The collaboration underscores AI's potential to address complex global issues, positioning OpenAI as a pivotal player in scientific and security advancements.

Speaker A [03:07]: "We're talking potentially groundbreaking stuff here."

---

4. Apple’s AI Strategy and Deepseek

Apple's approach to AI has traditionally been conservative, focusing on integrating AI within its tightly controlled ecosystem. However, recent developments indicate a nuanced shift in their strategy.

a. Tim Cook’s Comments on Deepseek

During a recent earnings call, Apple CEO Tim Cook made notable remarks about Deepseek, a Chinese AI startup gaining traction with its open-source models. Cook praised Deepseek's innovation, highlighting its potential to drive efficiency and impact profit margins.

Speaker B [03:55]: "He even called their models, quote, 'innovation that drives efficiency.'"

This endorsement suggests that Apple recognizes the disruptive potential of open-source AI models, potentially influencing their own AI development trajectory.

b. Apple's Hybrid AI Approach

Apple continues to maintain a hybrid approach by combining on-device AI processing with strategic partnerships, such as their collaboration with OpenAI. This strategy allows Apple to leverage external AI advancements while retaining substantial control over AI integration within their products.

Speaker B [04:35]: "They combine the power of on-device processing with strategic partnerships like the one they have with OpenAI."

However, the integration of third-party AI models raises questions about maintaining Apple's standard for privacy, security, and ecosystem cohesion.

c. Controversies Surrounding Deepseek

The episode also touches upon the controversies involving Deepseek, specifically accusations from OpenAI regarding unauthorized use of their models for training purposes. These allegations highlight ongoing disputes over intellectual property and the ethical use of AI technologies.

Speaker A [05:03]: "OpenAI has actually accused them of using their models for training... violating usage policies."

This situation underscores the complexities of AI development in a globally competitive environment, where intellectual property rights and ethical considerations are paramount.

---

5. Ethical and Societal Implications

The rapid advancements in AI, as discussed in the podcast, bring to the forefront several ethical and societal concerns:

a. Workforce Impact

Features like Google's 'Ask for Me' potentially automate tasks traditionally performed by humans, prompting discussions about job displacement and the need for workforce adaptation.

Speaker B [02:34]: "It's a great question... how do we adapt?"

b. Intellectual Property and Fairness

The controversy between OpenAI and Deepseek raises critical issues about intellectual property rights, data usage policies, and the fairness of competitive practices in AI development.

Speaker A [07:40]: "It's tricky... because AI often relies on these massive data sets for training."

c. Ethical AI Interactions

The deployment of AI features that interact with businesses and individuals necessitates stringent ethical guidelines to ensure transparency, respectfulness, and accountability.

Speaker B [08:19]: "How do we make sure that those interactions are ethical and respectful? How do we ensure transparency?"

d. AI Reliability and Trust

Apple's AI news summary feature faced setbacks due to generating misleading headlines and fabricating details, highlighting the ongoing challenges in ensuring AI reliability and maintaining public trust.

Speaker B [06:08]: "It started generating some misleading headlines... not a good look."

These discussions emphasize the importance of balancing innovation with ethical responsibility to foster a trustworthy and equitable AI-driven future.

---

6. Conclusion

The episode concludes with reflections on the swift evolution of AI technology and its pervasive impact across various sectors. Hosts A and B stress the importance of staying informed through reliable resources like AI Deep Dive to navigate the complexities of the AI landscape.

Speaker B [10:07]: "That's why resources like AI Deep Dive are so valuable."

They advocate for continuous learning, critical thinking, and active participation in shaping AI's trajectory to ensure that its development aligns with societal values and ethical standards.

Speaker A [11:10]: "Keep learning. Keep asking those tough questions... keep the conversation going."

The episode serves as a comprehensive overview of current AI advancements, strategic industry moves, and the pressing ethical considerations that accompany the technology's rapid growth.

---

Notable Quotes with Timestamps:

• [00:26] Speaker A: "You wouldn't even know it unless you like really dug into the changelog for their Gemini app."
• [01:04] Speaker B: "It offers better factuality... stronger performance for coding and math related tasks."
• [02:11] Speaker A: "So it just takes care of it."
• [02:34] Speaker B: "It's a great question... how do we adapt?"
• [03:55] Speaker B: "He even called their models, quote, 'innovation that drives efficiency.'"
• [05:03] Speaker A: "OpenAI has actually accused them of using their models for training... violating usage policies."
• [06:08] Speaker B: "It started generating some misleading headlines... not a good look."
• [08:19] Speaker B: "How do we make sure that those interactions are ethical and respectful? How do we ensure transparency?"
• [10:07] Speaker B: "That's why resources like AI Deep Dive are so valuable."
• [11:10] Speaker A: "Keep learning. Keep asking those tough questions... keep the conversation going."

---

Stay tuned to AI Deep Dive for more insightful discussions on the ever-evolving world of artificial intelligence.

Transcript

[A]

Foreign. Welcome back, everybody, to another Deep Dive. Today we're going to do kind of a whirlwind tour of all the top AI news stories. Yeah, we've been curating these from AI Deep Dive, which is always a great resource by the way. I think we should get right into it today, right?

[B]

Yeah, for sure. It really is a fascinating time to be following AI. You know, so much is happening so.

[A]

Quickly, so let's jump right in. First up, Google. Google, you know, the Big G, they've just come out with a new AI model, their flagship model. It's called Gemini 2.0 Pro experimental. But you know, the funny thing is you wouldn't even know it unless you like really dug into the changelog for their Gemini app.

[B]

Yeah, you know, I was thinking about that. It is kind of a surprisingly low key approach for a company like Google especially, you know, with all the pressure they're facing from Deepseek, you know, that Chinese AI startup, their open source models are, they're making waves.

[A]

So what do we know about this new Gemini upgrade?

[B]

Well, according to Google, it offers better factuality, which I think is really interesting. They're really focusing on the accuracy of their models and they're also saying it has stronger performance for coding and math related tasks, which, again, very important. But you know, there's a catch. Yeah, it's only available right now to a select group of users, the ones that have Gemini advanced, you know, their premium service.

[A]

So they're taking it slow.

[B]

Exactly, exactly. They're actually labeling it as an early preview and I think that's smart. They're acknowledging, you know, maybe it's not perfect yet.

[A]

Right, right. You don't want another public AI blunder, you know?

[B]

Exactly, exactly. Not in this competitive climate.

[A]

Not in this climate. Yeah. Well, speaking of practical applications, Google's also rolling out this new Ask for me feature.

[B]

Oh, yeah. This is really where AI starts to feel like something we would actually use every day, you know.

[A]

So tell me about that.

[B]

Okay, imagine you need an oil change, or you're trying to find a good mail salon, or maybe you need to schedule a doctor's appointment. And instead of you spending hours on the phone, you just tell Ask for me what you need.

[A]

Okay. So it just takes care of it.

[B]

Yeah, it contacts businesses, you know, on your behalf. It finds out about their services, their pricing, their availability and all that. Then it gives you like a nice summary.

[A]

Wow. I mean, that sounds amazing. It sounds like a huge time saver.

[B]

It is.

[A]

But you know, it does make you think what happens to the people whose jobs involve making those calls.

[B]

It's a great question, you know, it's a question we're going to be wrestling with as AI takes on more and more of these tasks. Yeah, I think we need to think about the impact on the workforce. You know, how do we adapt?

[A]

Absolutely.

[B]

But I mean, the technology itself is pretty impressive.

[A]

It is, it is. Let's move on to OpenAI. They've teamed up with the US National Laboratories and they're going to deploy their reasoning models on, you guessed it, the Venato supercomputer at Los Alamos National Laboratory. We're talking potentially groundbreaking stuff here.

[B]

Oh, absolutely, yeah. I mean, they're aiming to tackle some really big challenges. You know, disease treatment and prevention, cybersecurity, energy, innovation. I mean, they're even hoping to use AI to like deepen our understanding of the universe. Universe. And you know, another key area of focus for them is nuclear security. OpenAI could potentially play a major role in reducing those risks.

[A]

It is kind of mind boggling to think about the sheer scale of some of these projects. You know, we were just talking about AI, you know, calling nail salons, and now we're talking about AI potentially preventing nuclear war.

[B]

Right. I mean, it feels like it's touching every aspect of our lives.

[A]

It really does. Well, let's, let's bring Apple into the conversation. Tim Cook, Apple CEO, he made some interesting comments about Deepseek. Yeah, this is during a recent earnings call.

[B]

It's. Yeah, that's the Chinese company that's been making waves with its open source models, right? That's the one, yeah. You know, Apple, they're usually pretty tight lipped about their AI strategy for sure. But Cook actually went out of his way to praise deepseek. He even called their models, quote, innovation that drives efficiency. And he even talked about, you know, their potential impact on margins. You know that, that was really interesting to me.

[A]

Why is that?

[B]

It just suggests that even Apple, a company that's known for you know, its tightly controlled ecosystem, they're recognizing the potential of this open source AI. They're acknowledging that, you know, this new approach could be a game changer.

[A]

But Apple has traditionally taken kind of a different path, right?

[B]

They have, yeah, absolutely. They've been focusing on a hybrid approach. They combine, you know, on device processing with cloud based part partnerships like the work they do with OpenAI.

[A]

So it'll be interesting to see how this all plays out. Do they stick with their current strategy or do they start incorporating more of these third party AI models into their products?

[B]

It's a great question.

[A]

And speaking of Deepseek, they're not without controversy.

[B]

Yeah, that's true. OpenAI has actually accused them of using their models for training. They're saying it's potentially violating usage policies and all that.

[A]

Oh, wow. Okay.

[B]

Yeah, it raises some big questions about, you know, intellectual property rights in this age of AI.

[A]

Yeah, it sounds like it's not all sunshine and roses in the AI world.

[B]

No, it's, you know, there are definitely some gray areas here for sure and potential pitfalls we need to be aware of. But, you know, the innovation, it's undeniable.

[A]

Yeah. The innovation is happening so fast. Speaking of potential pitfalls, how about that Apple intelligence rollout? I feel like that was supposed to be a big deal and it's kind of.

[B]

Yeah, I mean, it was highly anticipated. They've been gradually introducing, you know, these AI powered features, but it hasn't exactly set the world on fire.

[A]

Right, right. And then you've got their, their AI news summary feature. They had to like pause that, right?

[B]

They did, yeah.

[A]

Why, what happened?

[B]

It started generating some misleading headlines.

[A]

Oh, wow.

[B]

There was even a case, I think, where it, you know, completely fabricated a detail in a BBC news story. Not, not a good look.

[A]

Yeah, I mean, that's not going to inspire confidence in their AI. Ouch.

[B]

Right. It just, it reminds us that we're still in the early stages for sure and there are going to be bumps along the way, but the progress is undeniable. And that's what I think makes it so fascinating to watch.

[A]

That's what keeps us coming back.

[B]

Exactly.

[A]

You know, it's amazing to think about all these stories and how they fit into this, you know, bigger picture of AI development. Like we've got American companies, Google, OpenAI, and then you've got this really ambitious deep Seq startup from China. And then Apple kind of like doing their own thing.

[B]

Yeah. It does feel like we're watching this high stakes chess match unfold. You know, each player kind of making these strategic moves, trying to, you know, outmaneuver the others.

[A]

And the prize is huge. Whoever dominates AI, I mean, they're going to reshape entire industries.

[B]

Absolutely. You know, even the global balance of power could be affected. Sure. But it's not just about, you know, technological dominance. Right. These companies are, they're walking this tightrope. They're trying to balance, you know, innovation with responsibility.

[A]

Yeah, yeah.

[B]

Like, you know, think about these allegations against Deep Seat, for example.

[A]

Right.

[B]

It really raises this whole issue of like, you know, intellectual property rights. In the AI world, you know, how do we protect the work of companies like OpenAI while still, you know, encouraging healthy competition?

[A]

It's tricky, right, because AI often relies on these massive data sets for training.

[B]

Right.

[A]

And sometimes those data sets can include, you know, copyrighted material.

[B]

Exactly, exactly. So the lines get blurred, you know, and then you add in these accusations of, you know, companies copying each other's work. It raises all these questions about fairness and transparency.

[A]

Yeah. I mean, if Deepseek really did use OpenAI's technology without permission, that really undermines the trust that's essential for this whole industry to thrive.

[B]

Precisely. Yeah. And this actually ties back to, you know, Google's Ask for me feature, as convenient as it sounds.

[A]

Right.

[B]

Think about it. We're essentially giving AI the power to interact with businesses and individuals, you know, on our behalf. So, I mean, how do we make sure that those interactions are ethical and respectful? How do we ensure transparency?

[A]

You know, it's like we're opening this Pandora's box of possibilities, but we need to be cautious about, like, the unintended consequences.

[B]

Exactly. And that brings us back to, you know, OpenAI and their collaboration with the national labs. Imagine the potential benefits, AI helping to cure diseases, you know, protect us from cyber attacks, even prevent nuclear war. But then it also raises the question of, you know, how do we prevent that same power from being misused?

[A]

It's a heavy responsibility.

[B]

It is.

[A]

It makes you think about, you know, Apple's more conscious approach. Yeah, they really seem to be prioritizing control over, you know, speed.

[B]

They are, and I think that's reflected in their hybrid approach. You know, they combine the power of, you know, on device processing with those strategic partnerships like the one they have with OpenAI. It allows them to maintain a level of oversight that you just don't get with a purely open source model.

[A]

But will that be enough to keep them competitive in the long run? I mean, as AI becomes, know, more and more integrated into our lives, will Apple's kind of closed ecosystem start to feel, I don't know, limiting?

[B]

Yeah, that's the billion dollar question. I mean, only time will tell. But I think for users, you know, people like you, the takeaway is this. Pay attention to these developments. You know, think critically about how AI is being used. Ask questions, demand transparency. Yeah, because ultimately the future of AI, it's going to be shaped by, you know, the choices that we make today.

[A]

One thing that really strikes me is just how fast this technology is evolving. It feels like every week there's like, you know, a new breakthrough, a new controversy, a new possibility. It's just. It's incredible.

[B]

It is. It's an exciting time to be following AI, but it can also be overwhelming. You know, that's why resources like AI Deep Dive are so valuable. You know, they help us make sense of this. This rapidly changing landscape.

[A]

Absolutely. And that's what we're trying to do here as well, you know, to break down these complex topics, you know, separate the hype from reality, and really empower you to make informed decisions about AI.

[B]

That's a great point. And that actually brings us to the end of our deep dive into the world of AI.

[A]

Wow. It feels like we just started.

[B]

I know, I know.

[A]

We've explored the latest developments, from those surprising launches to groundbreaking partnerships, and, you know, some really thorny ethical dilemmas. It's a lot to take in.

[B]

It is, but it's fascinating stuff.

[A]

It really is. And, you know, we've seen how AI is rapidly changing. You know, the way we live, the way we work, and, you know, just the way we interact with the world around us.

[B]

And I think the biggest takeaway for me is that the future of AI, it's not predetermined. You know, it's something we're actively shaping.

[A]

So, you know what? Keep learning. Keep asking those tough questions, and, you know, most importantly, keep the conversation going until next time. Thanks for diving deep with us.