The AI Daily Brief: Artificial Intelligence News and Analysis - How People Are Using AI for Health

Starting point is 00:00:00 Today on the AI Daily Brief, how people are using AI for health, and before that in the headlines, wouldn't be another year in AI without billions of dollars in fundraising. The AI Daily Brief is a daily podcast and video about the most important news and discussions in AI. All right, friends, quick notes before we dive in. First of all, thank you to today's sponsors. Optimizely, ZenCoder, robots and pencils, and super intelligent. To get an ad-free version of the show, which starts at three bucks a month, go to patreon.com slash AI Daily Brief, or you can subscribe on Apple Podcasts.

Starting point is 00:00:32 If you're interested in sponsoring the show, you can send us a note at sponsors at AIdailybrief.aI. And while you're at AIDailybrief.ai, you also might notice something interesting. Here in our list of options, there is a new option called AI Strategy Compass join the beta. Longtime listeners will have heard me talk frequently about super intelligent, a platform that helps companies figure out what AI to use and where. But now we're taking what we do in a hands-on and research-intensive way and turning it into a platform that is effectively a power tool for AI-strivate. strategists. So if you are someone who is responsible for AI transformation and AI adoption in your company, I want you to join this beta. I've only got a very small number of slots for this first wave. So again, if you are an AI strategist, anyone who's responsible for AI transformation and adoption,

Starting point is 00:01:17 go to AI Daily Brief.a.compan or just click on number five on AIdailybrief.ailyvriff. Fill out the form and we will be in touch very soon. Now with that out of the way, let's get to today's episode. Welcome back to the AI Daily Reef Headlines edition, all the daily AI news you need in around five minutes. We kick off today with a slate of fundraising stories. 2026 appears to be every bit as active as 2025 was when it comes to massive amounts of private capital flowing into AI startup leaders. The first news is that Anthropic is raising $10 billion at a $350 billion valuation. Now, Anthropic raising a bunch more money is not at all a surprise. Throughout the course of 2025, we saw their revenue accelerating at a rate even faster than OpenAI. Just yesterday, the entire episode was about how

Starting point is 00:02:03 enamored of Claude Code and Opus 4.5 everyone is. So a big investment round was always in the cards. At the same time, doubling your valuation from just four months ago, it is a serious sign of strength. According to the Wall Street Journal, the round will be led by Coatu and Singapore's sovereign wealth fund, GIC. CNBC confirms that a term sheet has been signed with the round expected to close in the coming weeks. Now, one of the interesting questions is what this suggests for IPO prospects this year. One of, not the only, but certainly one of the questions when it comes to whether Anthropic or Open AI will decide to go public this year is just how much capital remains available to them in the private markets. Now, I personally don't think that this is going to be a

Starting point is 00:02:41 major deciding factor because I think that the appetite for getting a piece of these companies, especially if they continue to grow the way that they've been growing, is basically unlimited. If anything, the only thing that I'm surprised about at this round is that they're only raising $10 billion. And I would expect that if Anthropics stays on the trajectory that they're on, that it will be just a very small handful of months before we see another round at an even higher valuation. That was not the only foundation model raised news this week. In fact, it wasn't even the biggest raise this week. XAI has announced that they've closed their Series E funding, raising a massive $20 billion. The fundraising has been ongoing for months, and although valuation was not

Starting point is 00:03:15 disclosed, reports suggest the post-money valuation is around a quarter of a trillion dollars. Once again, that would more than double the valuation established back in March of last year during the X merger. The deal was also upsized from the original $15 billion that XAI was seeking. Now, at this point, we're getting a little desensitized to the huge numbers in AI funding, so it is worth reinforcing that this is, in fact, one of the largest fundraising rounds in history. OpenAI's year-long soft bank-led fundraising round in 20-205 was the largest venture round in history at $40 billion. Besides that, the largest AI fundraising round we've had was Anthropics' $13 billion series F that closed last September. The investor list is also notable with longtime funder Valor

Starting point is 00:03:51 equity partners leading the round. Stepstone, Fidelity, the Cotter Investment Authority, MGX, and Barron Capital Group also all participated in the round. Invidia and Cisco participated as strategic investors, supporting XAI in quote, rapidly scaling our compute infrastructure and build out of the largest GPU clusters in the world. The fundraising announcement also had some big new claims about XAI's business. They said they ended 2025 with a million H-100 GPU equivalence, which is a big jump from the roughly half a million previously reported across the Colossus 1 and 2 data centers. The announcement also claimed 600 million monthly active users across the X and GROC apps. Now, this is a little difficult to compare to the Claude, ChatGBTGET, and Gemini numbers,

Starting point is 00:04:29 given that it's probably fair to assume that it skews towards X users that haven't really engaged with GROC, but it still certainly shows the distribution potential and the distribution potential is very large. If nothing else, the deal cements Elon's records as one of the most effective fundraisers in the history of Silicon Valley. Speaking on the XAI fundraising back at October, NVIDIA's CEO Jensen Huang said, the only regret I have is that I didn't give him more money. Almost everything Elon is a part of you really want to be a part of as well.

Starting point is 00:04:55 Now, one interesting thread about the state of the conversation comes from Greg Eisenberg, who tweeted about the rounds and said, Temperature Check, would you rather own stock in Anthropic at a $350 billion valuation, OpenAI at an $830 billion valuation, or XAI at a $230 billion valuation? And while the results are obviously going to be skewed towards the X audience, there are a lot of XAI and anthropic responses. Now, one more bit of fundraising news, which in any other world and any other period of time would be huge, but almost seems quaint compared to the numbers we were just discussing,

Starting point is 00:05:25 LM Arena has raised $150 million at a $1.7 billion valuation. This is the second venture round for the model benchmarking company since they decided to become a commercial startup. They previously raised $100 million seed round in May that saw the company valued at $600 million, making this round, of course, almost a 3x in less than nine months. Prior to last year, Elm Arena had been funded by a combination of grants and donations, operating as a research project in affiliation with UC Berkeley. Since its launch in 2023, the platform has grown exponentially to become a critical provider of model evaluations.

Starting point is 00:05:55 Users rate model outputs head-to-head across text, coding image, video, and several other modalities. The platform now sees more than 5 million monthly users across 150 countries and processes more than 60 million AI conversations per month. The resulting leaderboards have become so critical to the public's perception of model quality that some labs have been accused of gaming the system with specially tuned models to rise up the ranks. In September, Ellam Arena launched their first commercial service called AI Evaluation,

Starting point is 00:06:19 which allows enterprises, AI labs, and developers to hire Elam Arena to perform bespoke evaluations using their community of users. The company says the service has gone from zero to 30 million ARR over its first four months. Now, the takes on this one are wildly divergent. On the one hand, you have folks like AI Appreciator who says, Ella Marina is a poster child for the part of AI that's a bubble. $1.7 billion valuation? Crazy.

Starting point is 00:06:41 It's weak, low-taste signal if you want to measure model quality. Rather, it measures user preference over a narrow set of prompts that doesn't reflect real-world usage. On the other hand, you have Akash Gupta who writes, The headline here is 30 million ARR in four months, but I'm more interested in the business model underneath. Elam Arena built something that feels impossible, a crowdsourced evaluation platform that became the single biggest marketing lever in AI, and then figured out how to charge the labs using it. He points out that the 7.5 million in revenue that they've commanded so far is new revenue in a category that previously didn't exist. The real story, however, he says, is the flywheel they built.

Starting point is 00:07:14 35 million users show up to play a game. Two anonymous AI responses pick your favorite. Those users generate 60 million conversations per month. That data becomes the most trusted benchmark in the industry. Open AI, Google and XAI all need their models on that leaderboard, so they pay to get evaluated. The harder question he says is whether this holds. Ultimately, though, he points out, evaluation just became a billion dollar category. Now, it's not just private investors who are excited about AI right now. Google has surpassed Apple to become the second most valuable company in the world. Wednesday's market action saw Google stock rise by 2.5% to reach a $3.9 trillion market cap, overtaking Apple for the first time since 2019. The flip was emblematic of how the two tech giants

Starting point is 00:07:54 are navigating the AI era. Google used 2024 and 2025 to get fit, overhauling their AI org and releasing several groundbreaking models to catch up to the state of the art. Apple, on the other hand, suffered huge attrition in their AI staff, saw the resignation of their AI lead, and still haven't released the Apple intelligence features they showcased back in 2024. Looking ahead, the question will be whether Google can make a run at Nvidia to become the most valuable company in the world. They'll need another 18% gain to close that gap, but some analysts think that's likely. Canacorn analyst Maria Rips maintained her buy rating on a Wednesday note and lifted her price target to reflect a further 21% gain. While Google executed well in 2025, they still have a ton

Starting point is 00:08:30 of catalyst to roll out in this new year as well. They're launching their TPU chips as an external product for the first time, which presents challenges but also huge opportunities. But there's also other things like their Waymo division hitting an inflection point as AI-enabled self-driving cars reach maturation. Some folks, including me in my 2026 predictions, think that this year Google will catch up and overtake the juggernaut that is Nvidia. Today, Google surpassed Apple to become

Starting point is 00:08:52 the second biggest company in the world. Only a matter of time until they are number one. When, not if. If they stay on the trajectory they're on, I agree wholeheartedly, but for now, that is going to do it for today's AI Daily Reef headlines. Next up, the main episode. Most marketing teams aren't short on ideas,

Starting point is 00:09:09 but what they are short on is time. And that's exactly what optimally Opel gives you back, with AI agents that handle real marketing workflows. You know, like creating content and checking compliance, generating experiment variations, personalizing user experiences, analyzing pages for GEO, even tasks like approvals and reporting. It's your AI agent orchestration platform for marketing and digital teams, plugging seamlessly into the tools you already use, handling the boring busy work, and keeping everything on brand.

Starting point is 00:09:37 That leaves you marketers with more time to do your actual job. See what Opel can automate for your team by signing up for a free Enterprise Agentic AI workshop with Optimizely. Learn more at Optimizely.com slash the AI Daily Brief. If you're using AI to code, ask yourself, are you building software or are you just playing prompt roulette? We know that unstructured prompting works at first, but eventually it leads to AI slop and technical debt. Enter Zenflow.

Starting point is 00:10:04 Zenflow takes you from vibe coding to AI-first engineering. It's the first AI orchestration layer that brings discipline to the chaos. It transforms freeform prompting into spec-driven workflows and multi-agent verification, where agents actually cross-check each other to prevent drift. You can even command a fleet of parallel agents to implement features in fixed bugs simultaneously. We've seen teams accelerate delivery 2x to 10x. Stop gambling with prompts. Start orchestrating your AI.

Starting point is 00:10:31 Turn raw speed into reliable production-grade output at zenflow.3. Most companies don't struggle with ideas. They struggle with turning them into real AI systems. that deliver value. Robots and Pencils is a company built to close that gap. They design and deliver intelligent cloud-native systems powered by generative and agentic AI, with focus, speed, and clear outcomes. At Robots and pencils, engineers, strategists, designers, and applied AI specialists work together to move from idea to production without unnecessary friction. Powered by RoboWorks, their agentic acceleration platform,

Starting point is 00:11:02 teams deliver meaningful results including initial launches in as little as 45 days, depending on scope. If your organization is ready to move faster, reduce complexity, and and turn AI ambition into real results, robots and pencils is built for that moment. Start the conversation at robots and pencils.com slash AI Daily Brief. That's robots and pencils.com slash AI Daily Brief. Today's episode is brought to you by Super Intelligent.

Starting point is 00:11:25 Super Intelligent is a platform that very simply put is all about helping your company figure out how to use AI better. We deploy voice agents to interview people across your company, combine that with proprietary intelligence about what's working for other companies, and give you a set of recommendations around use cases, change management initiatives, that add up to an AI roadmap that can help you get value out of AI for your company. But now we want to empower the folks inside your team who are responsible for that transformation

Starting point is 00:11:48 with an even more direct platform. Our forthcoming AI Strategy Compass tool is ready to start to be tested. This is a power tool for anyone who is responsible for AI adoption or AI transformation inside their companies. It's going to allow you to do a lot of the things that we do at Superintelligent, but in a much more automated, self-managed way and with a totally different cost structure. If you are interested in checking it out, go to AID, Daily Brief.aI.compass. Fill out the form and we will be in touch soon.

Starting point is 00:12:15 Welcome back to the AI Daily Brief. ChatGBTGPT has introduced ChatGPT health, and this is both surprising and not surprising all at once. It's surprising in the sense that there has seemed at times in the past to be questions around how OpenAI wanted to handle people using ChatGPT in this way. It has appeared that the company has had concerns around people looking to ChatGPT as an alternative to their GP or doctor and the potential legal implications for that type of use, but clearly something is shifted and they are ready to lean in. Now, why it's not surprising is that this has seemed like a very big use case for some time, and recent data that the company shared made it clear just how big a use case it is.

Starting point is 00:12:57 Earlier in the week, OpenAI released a report called AI as a health care ally, how Americans are navigating the system with chat GPT. In that report, they share that over 40 million weekly active users globally prompt about health care every single day. One and four weekly active users prompt about health care each week, which for those doing the math is over 200 million users. Overall, more than 5% of all chat GPT messages globally are about health care. And so what are they asking? 55% are using chat GPT to check or explore symptoms, basically what we used to do with WebMD. 52% are using it to ask questions at any time of day. In other words, get fast information when you have a concern rather than having to wait for when

Starting point is 00:13:37 your doctor is available. 48% are using it to understand medical terms or instructions, given that medicine is in many cases a literal different language, and 44% are using it to learn about treatment options. OpenAI shares the truism that, as they put it, the health care system in the U.S. is a longstanding and worsening pain point for many. And outside of that high-level data, there's even more interesting indications of just how powerful and in what ways AI can be a healthcare assistant. Based on anonymized chat GPT message data, OpenAI found that nearly two million messages

Starting point is 00:14:07 a week focus on health insurance specifically. That includes everything from comparing plans and understanding prices to handling claims and billing. In underserved communities, users sent an average of nearly 600,000 health care-related messages each week. It also appears that outside of regular hours thing really matters as seven and ten health conversations in chat GPT happen outside of the normal hours that a clinic would be available. According to an open AI survey, three and five US adults report having used AI tools for health or health care issues in the past three months. and it actually really does seem like it's solving a problem. Digging deeper into this discovery that 600,000 health-related messages are sent

Starting point is 00:14:43 from users in underserved rural areas in the U.S., OpenAI writes, In the U.S., about one in five people live in rural areas, where populations skew older and face higher burdens of preventable disease and premature death. Inpatient care has become scarcer in rural communities since 2010, with 10 rural hospitals on average either closing or converting to models without inpatient beds each year. Financial pressure in these areas is widespread, nearly 46% of all rural hospitals operate with negative margins, and more than 400 across 38 states are considered vulnerable to closure. AI, they write, will not, on its own, reopen a shuttered

Starting point is 00:15:16 hospital, restore a discontinued OB unit, or replace other critical but vanishing services. But it can make a near-term contribution by helping people in underserved areas interpret information, prepare for care, and navigate gaps in access, while helping rare clinicians reclaim time and reduce burnout. Now, speaking of health care professionals, the report also talks about how they're using AI. They found that between 2023 and 2024, the percentage of American physicians who reported using AI for at least one use case jumped from 38 to 66 percent, and that survey data shows that 46 percent of U.S. nurses use AI at least once a week. The report closes with some ideas around how we could safely expand the use of AI in healthcare, including ideas like opening

Starting point is 00:15:54 and securely connecting the world's medical data to speed up scientific discovery, supporting workers' transitions into new healthcare professions that will be created and expanded by AI, and clarifying the regulatory pathway for AI medical devices for consumer use. So all of this came out on Monday of this week, pretty clearly setting up what happened next, which was the launch on Wednesday of ChatGPT Health. OpenAI's CEO of Applications, Fiji Simo wrote, The launch of ChatGPT Health is really personal for me. I know how hard it can be to navigate the healthcare system even with great care. AI can help patients and doctors with some of the biggest issues. She expanded on this in a larger blog post called What AI Can Do for a Broken System. She started with the personal story of being

Starting point is 00:16:33 in a hospital for a kidney stone and a secondary infection that had developed. After the resident in charge prescribed the usual antibiotic for that kind of infection, Fiji asked for a couple minutes before it was administered. She wrote, Because I've been dealing with a chronic illness for years, I had already uploaded a lot of my health records into Chachypt. I asked whether I should be taking this antibiotic, given my medical history, and Chachypte flagged that this particular antibiotic could reactivate a very serious infection I'd had a couple of years prior. She said that the resident was relieved and was glad Fiji had caught it, and when Fiji asked why it wasn't caught by someone at the hospital instead, the resident explained that she only has five minutes

Starting point is 00:17:07 per patient when making rounds, and health records aren't organized in a way that would make that sort of risk clear. Fiji goes on to argue that within the context of four key problems with the health care system today, AI potentially has a role to play in addressing each one. The first is the problem she had just described, which is doctors not having enough bandwidth. AI has the potential to absorb much more information and help medical professionals make better decisions. It can also help that medical information be more understandable for the patients. The second problem she writes is that the health care system is fragmented, but health requires looking at the full picture. Once again, potentially AI could be in the repository for a much

Starting point is 00:17:40 broader set of context. Then there's an issue of cost and access, which were explored extensively in the report that we just talked about. And finally, she writes, our health care model is reactive rather than preventative. Again, this is a known truism believed by many, regardless of political persuasion, but Fiji reminds, according to the CDC, five of the top 10 causes of death in the U.S. are associated with preventable and treatable chronic diseases. But our system is set up around going to see a doctor only when something is wrong. As some people have pointed out, we don't really have a health care system, we have a sick care system. Now Fiji argues with AI, anyone can have a daily companion to support their health journey.

Starting point is 00:18:15 So that's the why, but what is the what? What does chat GPT health actually include? First of all, it is a dedicated health experience inside the application. In the announcement post, they write, you can now securely connect medical records and wellness apps like Apple Health, function, and My Fitness pal, so ChatGPD can help you understand recent test results, prepare for appointments with your doctor, get advice on how to approach your diet and workout routine, or understand the trade-offs of different insurance options based on your health care patterns. The health section is going to be a dedicated space inside the application. They write that the conversations, connected apps,

Starting point is 00:18:45 and files associated with health are stored separately from all other chats, and that health actually has separate memory distinct from everything else. And while ChatGBT can use context from non-health chats to improve a health conversation, it cannot go the other way. Health information and memories do not flow back into non-health chats, and non-health conversations can't access files, conversations, or memories created within health. They also say that they built this in collaboration with more than 260 physicians who have practiced in 60 countries who have provided feedback on outputs over 600,000 times across 30 areas of focus.

Starting point is 00:19:17 Now, right now, this is not generally available. You have to sign up for a wait list, but people are already starting to weigh in. Simon Smith of Click Health wrote, I work in life sciences and chat GPT health is a big deal. One, it will be available to all users, 900 million plus people. Two, it will be secured for health use, encrypted and isolated.

Starting point is 00:19:34 Three, it will have key health app integrations and medical records. Four, it will be customizable with health-specific instructions. And five, trusted 260 doctors have given feedback 600,000 times. He wrote, I personally use chat GPT for health extensively. I have a health project and have uploaded loads of information into it,

Starting point is 00:19:50 including genetic information. I also update biometrics in that project monthly, and having direct integration with Apple Health will be awesome. Now, one of the things that people see most excited about is the ability to correlate data from different sources. When Simon wrote, question for Fiji-Simo or whoever is the product lead for chat GPT health, can I run custom analysis of Apple health data with charts? Like, can I ask a question like, how do my daily steps correlate with how well I sleep and get a chart that cross-references the data? Fiji said, yes, you can definitely do this and then actually gave an example of a coworker asking exactly that. Simon reposted and said,

Starting point is 00:20:22 I've tried to do biometric correlations in Apple Health, but it's a pain in the butt. Fiji replied to say yes, we'll be able to run correlations right in chat GPT health, very excited for this. And this correlation seems to be a big deal. Morgan Linton asks, curious if this can pull in data from Woop and Function Health and cross-reference. That's the main thing that's been missing for me is connecting blood test results with sleep exercise and heart rate data. Another reaction, of course, as with any lodge from OpenAI, is about all the startups that just became redundant. To leap Kumar writes,

Starting point is 00:20:49 I meet dozens of AI health startups every week and can tell you this is a big deal. Most of them will become redundant once this gets adoption. Your medical triaging, nutrition, fitness training, rehab, mental health all in one place. Or as Dr. Danish put it, 35 different digital health startups just died. Another threat of commentary is about how this creates a new type of defensibility. Akash Gupta writes,

Starting point is 00:21:09 this is a data mode play disguised as a feature launch. The numbers tell the story. He basically says if so many people are already using chat GPT for health, why build a special set of features, with the answer being, because right now those conversations are ephemeral. Users upload a lab result, get an answer, and the context evaporates. OpenAI can't train on it, can't personalize future responses, can't build compounding value, connect your EHR and Apple Health, now every conversation has continuity, your lipid panel from April, your statin prescription from May, your flu shot from last year, all indexed,

Starting point is 00:21:38 all searchable, all making the model more useful to use specifically. This creates a switching cost that's almost impossible to replicate. Google has search history. meta has social graphs. OpenAI is building the health graph. Simon Smith again points out, people on X have been crapping all over OpenAI for months, then chat ShpD health drops. Builds on and will expand high health use, will have huge beneficial impact and is hard to copy. Privacy, security, regulatory, not coming to rock or clot anytime soon. Critics seem oddly quiet. And yet, of course, it is not all that hard to find critics. One strand of criticism continues to be that Open AI is not focused. Journalist Shaquil Hashim writes,

Starting point is 00:22:14 My take from last summer that OpenAI is way too distracted to succeed continues to age beautifully. Timo Springer writes, it's no surprise that ChatGPT is struggling in losing market share. The product has become increasingly confusing for about 18 months. Features are launched and never touched again. Core workflows such as projects and Gptys perform significantly worse than competition. Newer features such as apps are brought to market half-baked. ChatGPT urgently needs a cleanup and improvement of its core product. AI entrepreneur Ethan Ding, however, doesn't agree.

Starting point is 00:22:41 He says this is a common take, but OpenAI's internal. strategy is remarkably sound. AI products are all or nothing. Good ones require no focus, see coding and chat GPT, no way to test without shipping to market, keep shipping till you hit the next jackpot, only jackpots can justify its valuation. Still, maybe the more obvious and potentially problematic critique is around privacy. Josh Long writes, so far zero response from anyone at OpenAI regarding who at the company can decrypt and view your health data and for what purposes. That's concerning, to say the least, and the vague claims about privacy and security in the blog post aren't helpful. Jonathan Shedler writes,

Starting point is 00:23:15 You guys do what you want, but I'm a health care provider, and there's no way I would upload my private medical data to an AI, and no way in hell I would upload mental health information. I am completely sympathetic to that point of view, and I'm sure that there will be many people who feel that way, but as someone who has spent a long time sitting around industries and projects that are convinced that people should care more about their privacy and data sovereignty

Starting point is 00:23:36 than they actually do, I don't think there's any chance that in general, those types of concerns will stop people who want answers about their health from just uploading an utter boatload of information to this system, even if others think they shouldn't. Now, obviously, I am someone for whom my major use cases around AI are all for work. However, this is one where I'm about to go sign up for this wait list.

Starting point is 00:23:56 I have vibe-coded numerous applications to try to tweak around highly specific health goals that I have, and I'm very excited to see it natively built into chat GPT. For now, that is going to do it for today's AI Daily Brief. Until next time, peace.

The AI Daily Brief: Artificial Intelligence News and Analysis - How People Are Using AI for Health

There aren't comments yet for this episode. Click on any sentence in the transcript to leave a comment.