6 Hail Mary's to Get Apple Back in the AI Game
21:52 Share

Today on the AI Daily Brief, six big moves that could get Apple back into the AI game. Before that in the headlines, Claude can now search the web. The AI Daily Brief is a daily podcast and video about the most important news and discussions in AI. To join the conversation, follow the Discord link in our show notes.

There was a time, friends, not so long ago, when many of the big foundation labs were concerned with giving their models web access. And this was from a safety perspective. They didn't like the idea of their models being able to go out there and learn on their own.

Well, now the lone holdout still operating under that paradigm has finally shifted because yes, after a long wait, Anthropic is finally adding web search to the Claude chatbot. Search has been one of the most requested features for Claude and is at this point completely de rigueur and standard for this competitive set. In their release blog post, Anthropic wrote, with web search, Claude has access to the latest events and information, boosting its accuracy on tasks that benefit from the most recent data.

And indeed, this is possibly the most obvious press statement I've ever read on the show, but it is also telling about just how much this feature set is absolutely integral to actual functionality of these tools. Indeed, plenty of people had in the interim built their own workarounds to try to give Claude access to the web. But at this point, web search is here, and a whole new set of use cases are opened up because of it.

Menlo's Didi Das writes, Now one note here is, of course, that as every company has launched their version of deep research, Claude was cut out because it didn't have these capabilities.

Some think this is a major competitive shift with, for example, the Archaeopteryx writing, with Claude now capable of web search, perplexity is cooked. I am not so sure about that, given that OpenAI has had web search for some time and perplexity remains a beloved product. And what's more, perplexity is not without resources to compete.

Bloomberg is reporting that Perplexity is in early talks to raise between $500 million and $1 billion in a round that would see their valuation double to $18 billion. The company's Series D closed in December, with the company taking in $500 million at a $9 billion valuation. Now, it's no surprise that the company has their pick of the litter from VCs to raise a bunch of money. But is there any actual new information in the reporting? We do have a source saying that Perplexity's current annual recurring revenue is near $100 million.

But this is still a pretty big cash pile. So what are they going to do with all that cash? Well, maybe we got a hint in this tweet from CEO Aravind Srinivas who wrote, Submit to this thread a list of things you found Manus AI to be really good at. We will make sure perplexity is as good or better. Research tool today becomes general assistant tomorrow.

Lastly today, Nvidia is looking to bring a sizable portion of their supply chain onshore to the U.S. Speaking with the Financial Times, CEO Jensen Huang said, Overall, we will procure over the course of the next four years probably half a trillion dollars worth of electronics in total, and I think we can easily see ourselves manufacturing several hundred billion of it here in the U.S.

These comments do seem like another tech company getting in line with President Trump's trade policy and the broader America First agenda. Huang did add, having the support of an administration who cares about the success of this industry and not allowing energy to be an obstacle is a phenomenal result for AI in the U.S. At the same time, this is a very strong endorsement of the efforts to set up domestic advanced chipmaking with the TSMC facility in Arizona.

That plant was designed to be able to produce cutting-edge chips, including the latest Blackwell architecture from NVIDIA. However, it was not a foregone conclusion that NVIDIA would proactively move a substantial part of the manufacturing onshore. There were, I think, two ways to view the Arizona facility. One, it being primarily a failsafe to ensure that earthquakes or war in Taiwan wouldn't bring the US AI industry to a screeching halt. And of course, there was also a supply chain control angle that bringing advanced chipmaking inside US borders allows greater control over chips for government and defense use.

Commenting on all this, Huang said, The most important thing is to be prepared. At this point, we know that we can manufacture in the U.S. We have a sufficiently diversified supply chain. He continued, If any disaster were to threaten production in Taiwan, Huang said, It will be uncomfortable, but it should be okay. Never a dull moment. However, that is going to do it for today's AI Daily Brief Headlines Edition. Next up, the main episode.

Hey, listeners, are you tasked with the safe deployment and use of trustworthy AI? KPMG has a first-of-its-kind AI Risk and Controls Guide, which provides a structured approach for organizations to begin identifying AI risks and design controls to mitigate threats. What makes KPMG's AI Risks and Controls Guide different is that it outlines practical control considerations to help businesses manage risks and accelerate value.

To learn more, go to www.kpmg.us slash AI Guide. That's www.kpmg.us slash AI Guide. Today's episode is brought to you by Vanta. Trust isn't just earned, it's demanded.

Whether you're a startup founder navigating your first audit or a seasoned security professional scaling your GRC program, proving your commitment to security has never been more critical or more complex. That's where Vanta comes in. Businesses use Vanta to establish trust by automating compliance needs across over 35 frameworks like SOC 2 and ISO 27001. Centralized security workflows complete questionnaires up to 5x faster and proactively manage vendor risk.

Vanta can help you start or scale up your security program by connecting you with auditors and experts to conduct your audit and set up your security program quickly. Plus, with automation and AI throughout the platform, Vanta gives you time back so you can focus on building your company. Join over 9,000 global companies like Atlassian, Quora, and Factory who use Vanta to manage risk and prove security in real time.

For a limited time, this audience gets $1,000 off Vanta at vanta.com slash NLW. That's V-A-N-T-A dot com slash NLW for $1,000 off.

Today's episode is brought to you by Super Intelligent and more specifically, Super's Agent Readiness Audits. If you've been listening for a while, you have probably heard me talk about this, but basically the idea of the Agent Readiness Audit is that this is a system that we've created to help you benchmark and map opportunities in your organizations where agents could

specifically help you solve your problems, create new opportunities in a way that, again, is completely customized to you. When you do one of these audits, what you're going to do is a voice-based agent interview where we work with some number of your leadership and employees to map what's going on inside the organization and to figure out where you are in your agent journey.

That's going to produce an agent readiness score that comes with a deep set of explanations, strength, weaknesses, key findings, and of course, a set of very specific recommendations that then we have the ability to help you go find the right partners to actually fulfill. So if you are looking for a way to jumpstart your agent strategy, send us an email at agent at besuper.ai, and let's get you plugged into the agentic era.

Today, we are continuing the saga of Apple, which is absolutely fumbling the pall so majestically when it comes to AI. And basically what we're going to do is we're going to cover the latest news, which is Apple shuffling around their leadership to try to get things headed in the right direction. But then, and you can probably guess how successful I think that this reshuffle will be based on the title of this episode, we're going to talk about a set of Hail Marys, big crazy plays that Apple could make that I think would have the potential to actually get them back into the AI fight.

Now, backing up and giving some context, although it's very clear that Apple's AI strategy has been in shambles for some time now, the problems really started coming to a head, at least in terms of public awareness last Friday.

AI Siri, or rather the lack of an AI Siri, was the focus of an all-hands meeting in Cupertino. Senior Director Robbie Walker hosted the meeting, as opposed to, for example, anyone involved in leading the AI team. He said that the Siri delays were ugly and embarrassing, those exact words, and he assured the team that, quote, intense personal accountability was being shared across leadership.

He stopped short of announcing the departure of any senior executives, but that's obviously where a lot of the speculation went. Now Bloomberg's managing editor and chief Apple correspondent reports that Tim Cook has lost confidence in AI head John Gianandrea. Bloomberg's Mark Gurman says that Vision Pro creator Mike Rockwell has been brought across to take over the Siri project. Rockwell will report to software chief Fred Federighi, cutting Gianandrea out of the project entirely. The decision was reportedly made at an off-site meeting of Apple's senior leadership held this week.

Now, honestly, the names don't really matter here. The TLDR is that the executive with the ownership over Siri has been removed and replaced by the executive who delivered the Vision Pro. Kerman writes, By tapping Rockwell, Apple is betting on an executive with proven technical experience. He has demonstrated the ability to ship new products and run an engineering organization with thousands of people. Rockwell has a knack for solving problems and often takes the role of evangelist for futuristic technologies.

Beyond this leadership change, though, by all accounts, there is a five-alarm fire going on at Apple. And while it would be tempting to write this off as a very weatherable storm for one of the most highly valued companies on Earth, I believe, and this might not surprise you, that Apple's failure to deliver on AI might be an existential problem. Now, before we get into my point about that, it's important to note that it is not just me. We talked last week about how John Gruber of the Daring Fireball blog had called for Tim Cook to step in and address the issue head-on.

He reflected on one of Steve Jobs' notoriously furious meetings when a product wasn't up to scratch and wrote, If such a meeting hasn't yet occurred or doesn't happen soon, then I fear that's all she wrote. The ride is over. Keep in mind that Gruber has been writing about Apple for over two decades with almost universal praise. And so when he says it might be over, it's a signal worth paying attention to.

And on top of all of this, Apple has been served with a lawsuit for false advertising regarding their Apple intelligence advertisements. The ads were pulled from YouTube last weekend in a final admission that none of the features they showed were coming anytime soon. Those ads were also a key topic at last week's all-hands meeting, with Walker placing the blame on the marketing department for wanting to showcase features that didn't properly function at the time. The lawsuit itself is whatever. Even if it gets settled, it's not going to make a dent in Apple's future. It's the fact that it exists, and

and that it's so clear and out in the open that there is just nothing behind Apple's curtain when it comes to AI. So let's talk about whether Apple has a way out of this. And I guess I'll start with this point around why it matters. In short, Apple is uncomfortably close to becoming Nokia. In the late 90s and early 2000s, Nokia seemed like an unstoppable force in the mobile handset market. 20 years later, they're a historical footnote. Although it is great that they're bringing back their brick phones for parents who don't want their kids lost in smartphone land.

Right now, the iPhone is Apple's most important product. Also right now, AI is features on a phone. In five years, though, AI is the phone, by which I mean AI dictates everything about our interactions with computers. AI eats hardware, user experience, and everything in between. And what that means is that people will leave Apple devices if those devices can't adapt.

This is not just about giving people a reason to upgrade to iPhone 16, which is what the business logic of Apple intelligence is presented as right now. It's that by the time that iPhone 20 arrives, there will very likely be a whole host of new form factor devices that are genuinely competing, and in many cases successfully competing with the phone as the control center for a new technological era. And indeed, that new era has a fundamentally different set of design principles that Apple is woefully out of sync with.

The design principles of the AI era are about being less controlled and more emergent. Experiences are more probabilistic than deterministic. Apple's strategy in the past has always been release it late but perfectly. But when it comes to the jagged frontier of AI, there is no perfect. There's no consumer design that's going to be perfected. There is just raw power, an absolute deluge of new capacity and capabilities.

You have to trust people to figure it out as they go and let experiences be messy as they do. This is unbelievably uncomfortable for a company like Apple, which has historically tried to control every single aspect of their experience. But I think in this new era, Apple is going to have to kill their darlings. There can be no sacred cows in Cupertino. So let's talk about the six Hail Marys that could get them into this game once again.

Here's an honorable mention. Apple needs new product leadership from outside the company. And this one isn't on the actual list because it's not a Hail Mary, it's an absolute must. As you'll see, much of what I'm going to recommend would require them operating not only differently, but in a way that is completely at odds with how Apple has succeeded in the past. It's going to involve knock-down, drag-out fights. And whoever is appointed to represent that disruptor view is

Tim Cook himself needs to empower that person with near total power to win most of those arguments. So with that, let's move on to the actual list. Number one with a bullet, Apple needs to kill Siri.

I don't mean that Apple needs to get out of the assistant game. I mean that Apple needs to destroy, bury, and plant some flowers on the grave of the Siri brand. The brand is at this point poison. Everything that is beloved about the Apple brand, the opposite is true for Siri. It has been a running joke for a decade. It represents the old way of doing things. It is a blaring reminder of Apple's failures. And Apple needs to kill it and put something new in its place. More specifically, they need to not wait to have something better...

They need to put someone else's stuff that is better right now on the iPhone as the assistant. It doesn't matter if it's OpenAI at the beginning. In fact, if you've ever interacted with OpenAI's voice recognition as compared to Siri's, you would absolutely kill to make that trade. The calculus is simple here. There is more harm to Apple's brand from Siri being an utter piece of junk than there is from throwing their hands up, saying we screwed it, and OpenAI is better, so here's OpenAI.

What I think that they're missing by not doing this is that Apple right now has platform lock-in with a ton of people. And that platform lock-in is not because it's such a pain to switch to Android or PC. The learning curve really isn't that high. Everyone's operating systems have gotten way better and more intuitive. The lock-in is because people genuinely like Apple experiences better. They like Apple hardware better. They like Apple's operating system better. Apple Silicon totally kicks butt. So why would you let software be the reason that people are forced to leave?

Yes, in the long run, owning the end-to-end experiences Apple always wants to do will allow them to make a better overall experience. But to get back there, they're going to have to concede a little bit along the way. Killing the Siri brand would instantly signal that Apple is playing to win by 2025's rules and realities, not 2015's rules. Number two, the second Hail Mary that Apple could go for, buying Anthropic.

Apple historically does not buy their way out of problems, so this would be another kill-your-darlings moment. So what would the logic for this be? If Apple wants to get back to owning the end-to-end experience, they need a foundation model, and frankly, at this point, they are absolutely, undisputedly too far behind to compete on that front. There's not enough compute to catch up. There's not enough talent to catch up.

They're going to have to partner on model access. They might as well just buy. OpenAI is too expensive and too messy. XAI is Elon's fiefdom. Mistral honestly isn't an insane option either. It would certainly be a lot cheaper than Anthropic. But Anthropic feels better suited to me for a couple of reasons. Anthropic is at the cutting edge of code generation. Code generation is going to be key for agentic controls, for allowing agents, in other words, to figure out how to interface with computers.

Agents interfacing with computers is perhaps the next big paradigm shift in user experience, and buying Anthropic gives them a leg there. The other reason is the model context protocol. MCP is becoming the default infrastructure for connecting AI and agents to data. It is currently racing ahead of all the other options and forcing companies like OpenAI to try to catch up with their agents' SDK. Now, nothing is won finally yet, but this is a core infrastructure play that could help Apple get back to owning the underlying experience.

Now, where it would get messy is that Anthropic's MCP strategy is clearly more Android than it is iOS, trying to get everyone using this open protocol. Knowing Apple, they'd probably want to figure out a way to restrict access to MCP. But still, all of this adds up to why buying Anthropic could be more interesting than perhaps it seems at first. Now, ability to buy Anthropic is a whole different story. Apple does have about as much cash on hand as Anthropic was valued at in its latest round, but I don't particularly get the impression that Anthropic is looking for a sale anytime soon.

Here's an alternative Hail Mary. Apple could buy DeepSeek and Manus. There's clearly something in the Chinese waters right now around not just models and efficiency and doing less with more, but around consumer experience. Remember, part of why DeepSeek exploded was that people just liked it better than the ChatGPT app. They liked the way that DeepSeek told you it was thinking. DeepSeek also had the right product instinct to give people access to the most performant model rather than limiting it by price and ability to pay.

Manus, likewise, wasn't necessarily innovating on the underlying model, but was wiring the agentic system together in ways that other people hadn't quite figured out, making it a much better, more performant experience on the way. So between the fact that these companies are moving blisteringly fast and they seem to have really, really good consumer product instincts, those are two attributes that are sorely needed in Cupertino right now. Honestly, this one isn't really even a Hail Mary. They should just try to do this as soon as humanly possible.

perhaps a little more Hail Mary-ish, is that Apple could decide to go all in on designing the agent operating system. One could argue that at the foundation of Apple's success in the modern era is the fact that people love their operating systems. Most consumers wouldn't think of it like that, but they're intuitive, easy to use, easy to figure out what you're trying to do. And right now, we are about to witness the birth of an entire new category of computer users, which are, of course, the agents. There are going to be agent-to-computer interactions, agent-to-person interactions,

and agent-to-agent interactions, all of which are going to require a totally different type of UI. Specifically, they're going to require new interfaces that aren't just tweaked versions of human interfaces, and this is once again an opportunity to own the foundation. It seems like it's something that Apple could do well, especially if they bought someone like Anthropic. You'll notice that a lot of these Hail Marys involve trading that big pile of cash for some new talent and new blood, and the fifth Hail Mary is no different. With this one, I'm suggesting buy 11 labs and

And straight up, and I mean, yes, ruthlessly copy meta Ray-Bans.

With the Ray-Bans, Meta is making a bet that pulling your phone out of your pocket every five seconds isn't necessarily people's best experience, and that there's going to be a lot of room for voice-mediated experiences in the future. The form factor of sunglasses, especially ones that people already like the design of, like the Ray-Ban, has worked really well. It's gotten meaningful consumer uptake. It's a popular product. And look, if nothing else, Apple makes great devices. They make great hardware.

To compete in this area well, though, they would need to also have access to really deep knowledge of and capacity in voice AI and audio generation. What I would argue here is that it doesn't matter that Meta is leading this right now, and it doesn't matter if people accuse Apple of following. There is tons that Apple can take from their learnings from the Vision Pro experience, and honestly, it would be great to see Apple acknowledge, which pretty much every normal human who thinks about it does, that

that augmented reality experiences in which we're still interacting with the real world are likely to represent a much bigger total addressable market than experiences that require us to go into some new world stuck on our couches with a giant headset. Now, that could be my own bias showing through, but when it comes to Apple's big plays in the AI space, this is one that actually seems really well-suited to them if they have the right talent. And speaking of really well-suited to Apple...

A final Hail Mary that they could take is buying figure and going all in on humanoid robotics. This path basically would accept a slow destruction of software and computing and reposition Apple as the command center for embodied AI. I think this is actually a lot less crazy than it seems. The innovation and disruption of the iPhone was understanding that mobile was going to represent this totally new paradigm, not just of computing, but of the way that people interacted with computers in the real world.

embodied AI, physical AI, robotics. This is a new generation of totally different types of computers interacting with the real world. Now, what makes this a little bit dicey is that Apple obviously could not pull off its car project, but they still have a lot of that latent talent there. It certainly needs new leadership. It needs people who aren't in the big tech nine-to-five hole and figure who's currently raising at $29 or $30 billion to

seems like a perfect fit in terms of both proven ability to execute, emergent leadership in the space, and the right price. The figure has also shifted off of their relationship from open AI, saying that when it comes to physical AI and embodied AI, you really need to own the experience end-to-end, which, bing, bing, bing, sounds a lot like Apple. Who knows, maybe even the leadership from a company like Figure could end up being that disruptor from outside that Apple so clearly and sorely needs. So with all of this said, do I think Apple is going to actually do any of this? I'd

I do not. I think that Tim Cook has been an incredible steward for a lot of Steve Jobs' legacy, but we are now officially in the period where the solid operator-manager-CEO is out of their depth and lost in the dogmas of the quiet past while the stormy present swirls around them, leaving them farther and farther behind.

I'm sure you guys are going to have a lot of strong feelings about this one, so I'm excited to see what you think. Does Apple have a chance? Am I overstating how far behind they are? Are there other strategies that you think are better than the ones I've suggested here? Share your thoughts in the comments, and we will keep an eye on the situation. For now, though, that is going to do it for today's AI Daily Brief. Until next time, peace.