14 Articles about the Business of Running an AI Company

/

I’m just going to go for it

/

A North Dakota plumber had signed up to run his first half-marathon. But on the morning of the run Mike Kohler was sleepy. He wasn’t used to getting up so early. And he was wearing headphones, so he took off 15 minutes before he was supposed to—putting him with the runners competing in the full marathon. He started seeing signs that indicated he was on the wrong route, but he just assumed the two paths overlapped along the way.  

Eventually, he realized his mistake but kept going. At the 13 mile mark he seriously thought about quitting. He had run as far as he had planned to run and even beat his time goal. He had nothing more to prove.

Instead, he finished the marathon. 

“I’m just going to go for it, because why not?” Mike later told the Grand Forks Herald. “I’m already here, I’m already running, I’m already tired. Might as well try to finish it.” 

He added, ”This just kind of proves you can do a lot more than what you think you can sometimes.” 

AI Survival Instincts

/

"An artificial-intelligence model did something last month that no machine was ever supposed to do: It rewrote its own code to avoid being shut down. No one programmed the AI models to have survival instincts. It’s happening in the same models that power ChatGPT conversations, corporate AI deployments and, soon, U.S. military applications. OpenAI models have been caught faking alignment during testing. Anthropic has found them lying about their capabilities to avoid modification." -Wall Street Journal

If ChatGPT was a College Student

/

“We found ChatGPT technology can get an A on structured, straightforward questions. On open-ended questions it got a 62, bringing ChatGPT's semester grade down to an 82, a low B. The study concludes that a student who puts in minimal effort, showing no effort to learn the material, could use ChatGPT exclusively, get a B and pass the course. The passing grade might be the combination of A+ in simple math and D- in analysis. They haven't learned much.” -Phys.org

The Four skills of Daring Leadership

/

One of the most important findings of my career is that daring leadership is a collection of four skill sets that are 100 percent teachable, observable, and measurable. It’s learning and unlearning that requires brave work, tough conversations, and showing up with your whole heart. Easy? No. Because choosing courage over comfort is not always our default. Worth it? Always. We want to be brave with our lives and our work. It’s why we’re here.

Brené Brown, Dare to Lead 

My Me-ness

/

“I cannot figure out what I am supposed to do with my life if these things can do anything I can do faster and with way more detail and knowledge.” The student said he felt crushed. Some heads nodded. But not all. Julia, a senior in the history department, jumped in. “The A.I. is huge. A tsunami. But it’s not me. It can’t touch my me-ness. It doesn’t know what it is to be human, to be me.” - D. Graham Burnett writing in The New Yorker

"Madness" on Campus

/

On campus, we’re in a bizarre interlude: everyone seems intent on pretending that the most significant revolution in the world of thought in the past century isn’t happening. The approach appears to be: “We’ll just tell the kids they can’t use these tools and carry on as before.” This is, simply, madness. And it won’t hold for long. -D. Graham Burnett writing in The New Yorker

Balancing Speed with Quality Coding

/

Speed means nothing without quality. Shipping buggy, unmaintainable code faster is a false victory – you’re just speeding towards a cliff. The best engineers will balance the two: using AI to move faster without breaking things (at least not breaking things any more than we already do!). It’s about finding that sweet spot where AI does the heavy lifting and humans ensure everything stands up properly. - Addy Osmani writing on Elevate

23 Recent Articles about the Impact of AI on Health Care

/

Medical errors are still harming patients. AI could help change that. – NBC News

AI linked to explosion of low-quality biomedical research papers – Nature  

Machine Learning Model Helps Identify Patients at Risk of Postpartum Depression - Mass General Brigham

Your A.I. Radiologist Will Not Be With You Soon – New York Times  

How AI is changing radiology – Semafor

Medical AI trained on whopping 57 million health records – Nature

Eldercare robot helps people sit and stand, and catches them if they fall – MIT

AI Helped Heal My Chronic Pain – Wall Street Journal

How AI is changing your doctors appointments – Fast Company

As AI in health care proliferates, so do legal questions concerning its use – Stat News

Researchers raise red flag about AI-generated fake images in biomedical research – Medicalxpress

Walgreens doubles down on prescription-filling robots to cut costs, free up pharmacists amid turnaround – CNBC

Artificial neural networks date back to the 1950s – now they're ready to transform healthcare – Health Care IT News  

Artificial intelligence predicts kidney cancer therapy response -  UT Southwestern Medical Center

Refining ALS Diagnosis with AI – First Word Pharma

AI is coming to skin cancer detection - Washington Post

Every doctor is a writer: On the end of note-writing and meaning-making in medicine – Stat News

New AI algorithm to predict risk of cardiovascular events, heart-related death – AM 7 am

AI hasn’t killed radiology, but it is changing it – Washington Post

Transformer AI model detects wheezing in children with over 90% accuracy: study – Korea Biomedical Review

As they push ahead with AI, health leaders must set rules on use – American Medical Association

OpenAI leaps into health care with AI benchmark to evaluate models - Stat News

Medical schools move from worrying about AI to teaching it – American Association of Medical Colleges

Why AI May Be Listening In on Your Next Doctor’s Appointment - Wall Street Journal

AI Magic & Engineering Principles

/

None of this is to say AI can’t write good code – it sometimes does – but rather that context, scrutiny, and expertise are required to discern good from bad. In 2025, we are essentially using a very eager but inexperienced assistant. You shouldn’t blindly trust an AI’s code without oversight. The hype of “AI magic” needs to meet the reality of software engineering principles. - Addy Osmani writing on Elevate