- Everyday AI
- Posts
- Putting patients first with medical AI
Putting patients first with medical AI
Microsoft may be replacing OpenAI to power Copilot, xAI invests in another AI gigafactory, Palantir and U.S. Army team up on AI.
Everyday AI
March 07, 2025
š Subscribe Here | š£ Hire Us To Speak | š¤ Partner with Us | š¤ Grow with GenAI
Outsmart The Future
Today in Everyday AI
7 minute read
š Daily Podcast Episode: Why do clinical trials take a decade and cost millions of dollars? Well, that story will change with GenAI in the mix. Give it a listen.
šµļøāāļø Fresh Finds: An AI tool that helps you build AI tools, Celine Dion gets ripped using AI, Apple confirms AI delays and more. Read on for Fresh Finds.
š Byte Sized Daily AI News: Microsoft may be replacing OpenAI to power Copilot, xAI invests in another AI gigafactory, Palantir and U.S. Army team up on AI. For that and more, scroll on to read more in Byte Sized News.
š§ Learn & Leveraging AI: How is the intersection of LLMs and Medical research putting patients first? Weāll break it down. Keep reading for that!
ā©ļø Donāt miss out: Did you miss our last newsletter? We talked about Microsoft going after Salesforce with new AI sales agents, Anthropicās AI policy suggestion for White House, Alibaba unveiled QwQ-32B and more! Check it here!
Top Reason For AI Failure: Cognitive Bias š§
(Almost) Everyone hates the medical system. It's slow. It's expensive. It's archaic. GenAI is starting to change that. Find out how from an industry leader.
Join the conversation and ask Jordan and Lina questions on medical AI here.
Also on the pod today:
ā¢ Why medical trails take a decade ā±ļø
ā¢ How AI speeds up medical research š§¬
ā¢ How medical AI is becoming like engineering š©āš¬
Itāll be worth your 35 minutes:
Listen on our site:
Click to listen
Subscribe and listen on your favorite podcast platform
Listen on:
Hereās our favorite AI finds from across the web:
New AI Tool Spotlight ā Toolable helps you build and share your own AI tools, Hedra takes your AI content game to the next level, Fynix helps you code faster with AI.
AI in Media ā No, thatās not a new Celine Dion song. Thatās AI.
Text to Speech ā Google Cloud and EleveLabs teamed up for enterprise text-to-speech
AI Regulation ā Google doesn't have to sell off its AI division, as U.S. antitrust case is dropped.
Apple Intelligence ā Apple confirmed that its full AI-powered Siri is delayed.
AI in Government ā The U.S. government is reportedly using AI to detect āPro Hamasā social media posts and revoking student visas.
AI in Society ā This California mom says her closest relationships are with AI chatbots.
1. Microsoft Exploring New AI Allies Amid OpenAI Rift š
As Microsoft reportedly moves away from OpenAI, itās testing AI models from xAI (Elon Muskās venture), Meta, and DeepSeek, according to The Information. This signals Microsoftās effort to diversify its AI sources for Copilot, its workplace assistant, as it looks to improve performance and cut reliance on ChatGPT.
These potential partnerships could reshape the AI landscape, with Microsoft hedging its bets across emerging competitors like DeepSeekāan open-source model from Chinaāthat could challenge existing players.
2. Palantir Deploys AI-Powered Titan Systems to U.S. Army šŖ
Palantir Technologies is delivering its first AI-enabled Tactical Intelligence Targeting Access Node (TITAN) systems to the U.S. Army, marking a major step in battlefield tech evolution, according to CNBC.
These mobile ground stations combine AI with space sensor data to boost strike precision and strategyāall packed into trucks for on-the-go intelligence. The $178 million contract positions Palantir as the first software company leading a substantial hardware defense program, signaling a shift in military reliance on AI.
3. xAI Expands in Memphis with Massive AI Data Center Deal šļø
Elon Muskās xAI has secured a 1 million-square-foot property in Southwest Memphis to boost its AI data center capacity, according to the Memphis Chamber of Commerce.
This expansion will complement its existing Memphis facility, Colossus, which xAI plans to supercharge with 1 million Nvidia GPUs this yearā10x last yearās count. The move comes as the company reportedly eyes a $10 billion fundraising round to fuel its AI ambitions, following a $5 billion GPU server deal with Dell.
4. Microsoft Investing More in its own Reasoning Model š¤
Microsoft is reportedly developing its own advanced AI models, codenamed MAI, as it looks to reduce reliance on OpenAI for powering tools like Microsoft 365 Copilot, according to The Information. These in-house models aim to rival OpenAIās GPT-4 in performance and could hit the market later this year as APIs for developers.
Mustafa Suleymanās team is testing the MAI models and experimenting with "reasoning" techniques to solve complex problems, signaling a potential shift in Microsoft's AI strategy.
5. Reflection AI Emerges with $130M to Chase Superintelligence š°
Reflection AI, a startup founded by ex-Google DeepMind researchers Misha Laskin and Ioannis Antonoglou, has launched with a hefty $130 million in funding.
Backed by heavyweights like Sequoia Capital, Nvidia, and LinkedInās Reid Hoffman, the company is valued at $555 million and aims to develop "superintelligence" ā AI capable of handling most computer-based work. Their first step? Building autonomous AI agents to tackle coding tasks like bug detection, memory optimization, and software testing.
š¦¾How You Can Leverage:
Robots cranking out 2 MILLION experiments weekly.
Scientists saving 60% of research time.
And patients who would've been TOAST in 2019 are now alive and kicking.
All because of Large Language Models completely transforming how medical research gets done.
And ultimately, that means healthcare facilities throughout the world can start putting patients first.
That's what Lina Nielsen, SVP and Head of Platform and Product at Recursion, shared on todayās show.
Lina took us behind the scenes and showed how AI is demolishing the $2.5 billion, decade-long drug discovery process.
Hereās what ya need to know. š
1. The million-dollar insights hiding in your failed experiments š¤¦
Traditional drug discovery is brutally inefficient.
$2.5 billion and a decade to create ONE medicine.
Lina explained why: when researchers hit dead ends, valuable data gets archived in notebooks and forgotten forever.
Recursion flipped this model.
They built a system where "rather than starting over when you hit a dead end, that data can be compared and joined together quantitatively over time."
The result? Patients who would have died two years ago are alive today because their AI found treatments hidden in abandoned experimental data.
This isn't marginally better drug discovery. It's a fundamentally different approach to scientific knowledge.
Try This
Create a "failure archive" that treats dead-end projects as future assets, not embarrassments.
Explore how companies like Edison Research and WD-40 built success by systematically mining their failures.
Run a quick workshop where teams revisit three abandoned projects looking specifically for reusable components.
Challenge: Find ONE valuable idea in your company's "digital graveyard" this month.
2. You're running AI backward šāā”ļø
Everyone's using the same AI playbook:
Collect massive data
Train models
Pray for insights
Lina revealed Recursion is doing the exact oppositeāand winning.
They're creating comprehensive "world models" of biology first, then running targeted experiments only to validate the most promising predictions.
The impact? Their scientists spend 60% less time reviewing literature before advancing drug candidates.
This isn't just for pharma companies. Every industry with complex systems is about to undergo this exact transformation.
Try This
Explore how NVIDIA and Google DeepMind are using "world models" in completely different domains.
Google is looking to use its Genie 2 world model for future general agents, while NVIDIA has its Cosmos World Foundation Model to accelerate physical AI progress.
3. Biology is becoming software engineering š©āš¬
Ten years ago, biologists quantified results with literal plus signs in research papers.
(Not kidding. + = a little. +++ = a lot.)
Today?
Precise analytics provide exact measurements.
Lina explained this isn't just better toolsāit's biology transforming from observation to engineering.
NVIDIA CEO Jensen Huang predicted this exact shift. Lina confirmed it's happening now.
The catalyst?
CRISPR technology. A decade ago, we could read the genome but not edit it effectively. Today, CRISPR enables precise editing, creating exponential data for machine learning.
This patternāidentifying unique data sources in your field, then applying MLāwill repeat across every industry.
Try This
Research how quantification transformed other industries before yours.
How did manufacturing change when subjective quality became precise measurement?
Identify one business process where your team still relies on "gut feel" instead of data.
Create a simple 1-5 scale for ONE previously unmeasured variable in your customer experience.
Challenge: Find something your industry considers "unmeasurable" and design an experiment to measure it anyway.
ā
Numbers to watch
$10 Billion
Anysphere, an AI startup, in talks to fundraise at a whopping $10 billion valuation.
Now This ā¦
Let us know your thoughts!
Vote to see live results
If our newsletter was sent at a more consistent time, would you read it more?Be honest. Our feelings won't be hurt |
Reply