Apple stock could be a winner in the AI ​​race

my part of resistance It was a series of cartoon giraffes bathing in a toilet. I suggested to her wife that it might be good for the bathroom wall. (She wasn’t that excited.)

This is amazing technology, but a complete waste of time.

Meta’s new chatbot is based on a modern open source large-scale language model called Llama 3, and has all the features you need for an artificial intelligence chatbot in 2024. I have a recommended list of albums and a bulleted list of comparison data for sport utility vehicles. But I’m not ready to update my view on the current state of generative AI. As an everyday consumer tool, I’m still not impressed. (My colleague Eric J. Savitz takes a different view.)

Still, I try to keep an open mind. The stock market has leveraged AI to create trillions of dollars in new value. It must be worth something. I’ll get to the “Wall Street” portion of this column, but first I’ll detail my exploration of AI valuation.

Shortly after Meta.ai’s lark, I spoke with Dr. Peter Noseworthy, professor of cardiology and medical director of business development at Mayo Clinic. I reached out to Mayo because I wanted to go beyond chatbot chatter.

It didn’t take long. Within minutes, Noseworthy spoke to Mayo about his AI projects he was working on and the life-saving impact of clinical trials. He combined technology, business, and medicine in a way I had never heard of before.

“One of the problems with health care is that it is delivered as a service and as a result is not scalable,” Noseworthy told me. Perhaps there are aspects of healthcare that can be transformed from services to products using large language models and AI. And once that happens, it’s scalable, it’s portable, and people can carry these things in their pockets. ”

He said Mayo’s team is using AI to find markers of disease in commonly performed electrocardiograms. “It turns out that the electrocardiogram is actually a very powerful marker for all kinds of diseases that aren’t dependent on the ECG, even for people who make a living reading ECGs.”

Eventually, he says, these ECG tools will also be available on devices such as the Apple Watch. “Then the consumer wears a clinical-grade His ECG tool on their wrist to deploy our His AI.”

We talked about phones as a ubiquitous form factor. “We can detect all kinds of signals about health and disease in patient voices, especially if trends change over time,” Noseworthy says. Similar signals could be obtained by a phone’s pedometer or by measuring changes in the user’s gait.

He likened AI and the phone to a mother quickly spotting a disease on her child’s face. “This kind of clinical gestalt from an astute observer is probably reproducible with AI. And we record our faces dozens, if not hundreds, of times every day. There may be opportunities to take advantage of that.”

Mobile phones have already taken over many roles in our lives, but doting mothers will take on a new role.

Noseworthy also provided us with another framework for considering chatbots. “The closest thing to a patient-doctor interaction is probably a generative AI LLM solution in the form of a medical chatbot,” he says. “If you had asked me six months ago, I would have thought that what I believed in was years away.Right now, he’s only six months away from being clinically useful. I think it might take a month.”

Noseworthy said Mayo is actively working on these chatbots. “It will take several more years to fully emulate clinical acumen,” he added.

In the short term, chatbots could help patients adjust their medications or ask questions about drug interactions, Noseworthy said.

“The most mature version of the technology will allow patients to talk to a chatbot in natural language to discuss their symptoms and concerns and receive high-quality, vetted medical information presented in an easy-to-understand and natural conversation. “I’ll be able to do that,” he told me.

These kinds of personalized tools take AI to another level.

I spoke to the founder and CEO of a company called Limitless. The company says it is using AI to give people the superpower of memory. I was already testing an early product from a company called Rewind.ai. This product uses text and voice recognition to record everything that comes from your Mac’s screen and speakers, including calls, texts, and emails. The app can then use large-scale language models to answer questions about our lives. “Tell me about your last conversation with your mother.” What time is our lunch? ”

The company’s new product, also called Limitless, offers apps for Mac, Windows, and the web. Last week, the company launched a wearable pendant with a microphone that users can attach to their clothing to record actual conversations without a device.

All of this presents great potential and, of course, significant privacy concerns. Limitless founder Dan Siroker says the two are not contradictory. “It starts with the basic premise that privacy should not come at the expense of convenience,” he says. “Facebook and all ad-driven companies have done a tremendous amount of damage to the world by making us think it should be an option – that we have to sacrifice privacy for convenience.”

The new pendant offers a consent mode that only tracks audio that you have given permission to record. The company says 10,000 orders have already been placed for the $99 pendant, which is expected to ship in August. Limitless says all data brought into the cloud is encrypted with a private key and cannot be accessed by anyone, including the company. Data submitted through the Large Language Model for summarization and transcription will be deleted after 30 days.

Limitless has raised $33 million from venture capital firms including Andreessen Horowitz and New Enterprise Associates, as well as OpenAI founder Sam Altman.

The key to Limitless is providing context to the AI ​​models, Siroker said. This would be like providing a prompt to a chatbot that says, “I like the colors blue and green.” Choose the best benjamin moore paint colors. ”

However, at Limitless, we’re about more than just a love of blue and green. Every conversation you’ve ever had can be sent. Eventually, AI models trained on common internet data will bring personalized answers to our lives.

So what does this mean for stocks? AI’s potential is already driving a massive market rally.Ironically, Apple
,

Companies that know more about our lives than any other company are being left behind.

Consider the Apple Health app on my phone. The app stores years of data on my daily steps, heart rate, number of times I stand per day, exposure to loud noises, blood oxygen levels, and more. There are thousands of photos with location and timestamps in a few rows on the home screen. Then there’s my text. With my permission, Apple could feed all of this into a large-scale language model to create an absurdly detailed diary of my life. This is a self-help book about steroids.

Yes, there are privacy concerns, but all this data is already provided to Apple. I have no problem with the company using it more personally.

The current blow to Apple is different from Big Tech rivals Meta and Alphabet.
,

Or startup OpenAI, which has yet to create its own large-scale language models.

I have no proprietary information as to whether that will be the case. CEO Tim Cook said Apple is spending a lot of time on AI, and more details will be announced later this year. The company will likely make a major announcement at its annual developer conference in June. But for now, investors have largely given up on Apple and AI. And that feels like a huge oversight.

Late last month, Ben Reitzes, an analyst at Melius Research, published a note titled “Apple May Still Have the Last Laugh” about the Cupertino, Calif., company’s AI efforts.

Mr. Reitz has covered Apple for more than 20 years as an analyst and banker. “The reason we wrote this memo was to say, remember, these are market makers. And maybe we should listen to them,” Reitz told me. said. “Even going back to the original example when Steve Jobs was trying to build his iTunes. He didn’t make music. He made music more fun.”

Reitzes says Apple doesn’t need to develop world-beating AI models to be a winner in this new world. “It doesn’t have to be Microsoft.”
.

It doesn’t have to be OpenAI or Gemini. Apple needs to be Apple, and that’s providing an interface to enjoy where apps are going in this new world. We think they can do it, and they are carefully starting to lay the groundwork for it, with major construction probably starting in June. ”

So far, Apple is down 14% this year, while Apple is up 4%.

S&P500

AI leader Nvidia rises 56%
.

Melius said there are legitimate concerns about Apple’s iPhone sales in China, where the company appears to be rapidly losing market share. And I understand why some investors are concerned about the company’s past AI efforts.

But Apple has been here before. From September 2012 to September 2013, until the iPhone 5 and iPhone 5S were released, Apple stock fell 33%, while the S&P 500 index rose 18%. At that time, Samsung Electronics was gaining momentum with large-screen mobile phones.

Then, at the end of 2014, Apple released its first large-screen phone, the iPhone 6. Since then, Apple stock has returned 635%, while the S&P 500 has returned 198%.

To be sure, switching to AI is more complicated than building a phone with a larger screen. However, Apple’s dominance remains the same. And it is at the center of all change, whether real or artificial.

Write destination Alex Yule (alex.eule@barrons.com)