Scott McGrath
@smcgrath.phd
Biomedical Informatics PhD • CITRIS Health @UC Berkeley • FAMIA • Focusing on Informatics and AI in medicine • Linfield U. Grad • Missoula MT
https://citris-uc.org/people/person/scott-mcgrath/
https://citris-uc.org/people/person/scott-mcgrath/
Pinned
Scott McGrath
@smcgrath.phd
· Sep 22
Decided to try something new out. If you like the ML Feed: Trending, try out the new ML Feed: Blend!
Instead of just showing the trending ML posts from the past 24 hours, this will show new posts mixed in with popular posts. It should provide a little more variety!
Instead of just showing the trending ML posts from the past 24 hours, this will show new posts mixed in with popular posts. It should provide a little more variety!
This one is beyond bizarre.
Proving Silicon Valley is beyond satire, a YC company launched "Chad: The Brainrot IDE." It integrates TikTok and Tinder into the coding window, claiming to boost productivity by preventing "context switching" while AI tasks run.
#MLSky
Proving Silicon Valley is beyond satire, a YC company launched "Chad: The Brainrot IDE." It integrates TikTok and Tinder into the coding window, claiming to boost productivity by preventing "context switching" while AI tasks run.
#MLSky
'Chad: The Brainrot IDE’ is a new Y Combinator-backed product so wild, people thought it was fake | TechCrunch
Chad: The Brainrot IDE is an actual product that pairs vibe coding with brainrot activities like gambling, Tinder and games.
techcrunch.com
November 13, 2025 at 4:06 PM
This one is beyond bizarre.
Proving Silicon Valley is beyond satire, a YC company launched "Chad: The Brainrot IDE." It integrates TikTok and Tinder into the coding window, claiming to boost productivity by preventing "context switching" while AI tasks run.
#MLSky
Proving Silicon Valley is beyond satire, a YC company launched "Chad: The Brainrot IDE." It integrates TikTok and Tinder into the coding window, claiming to boost productivity by preventing "context switching" while AI tasks run.
#MLSky
The rise of AI companions introduces a new wrinkle in family law. Attorneys report chatbot romances are increasingly cited in divorce proceedings, raising complex questions about infidelity, misconduct, and the dissipation of marital assets.
#LawSky #MLSky
#LawSky #MLSky
AI Relationships Are on the Rise. A Divorce Boom Could Be Next
Secret chatbot flings are creating new legal challenges for married couples when it comes to infidelity.
www.wired.com
November 13, 2025 at 3:51 PM
OpenAI is exploring consumer health, a sector where other tech giants have struggled. By tapping into ChatGPT's massive user base, the company is weighing tools like personal health assistants and data aggregators.
#MedSky
#MedSky
OpenAI is weighing a move into consumer health apps
Healthcare investors think OpenAI could build a personal health assistant complete with medical records, a project that's challenged Big Tech.
www.businessinsider.com
November 12, 2025 at 6:36 PM
OpenAI is exploring consumer health, a sector where other tech giants have struggled. By tapping into ChatGPT's massive user base, the company is weighing tools like personal health assistants and data aggregators.
#MedSky
#MedSky
In Trump's second term, clemency has largely bypassed the formal DOJ process, favoring allies. While 10,000 petitions wait, pardons are secured via back channels. Only 10 of the roughly 1,600 grantees had filed through the official clemency office.
How Trump Has Exploited Pardons and Clemency to Reward Allies and Supporters
The president’s second term has brought a flood of clemency for allies and donors — as well as felons, like him, who were convicted of financial wrongdoing.
www.propublica.org
November 12, 2025 at 4:19 PM
In Trump's second term, clemency has largely bypassed the formal DOJ process, favoring allies. While 10,000 petitions wait, pardons are secured via back channels. Only 10 of the roughly 1,600 grantees had filed through the official clemency office.
Goodfire.ai researchers isolated memorization from reasoning in neural networks. They found they could surgically remove verbatim recall while preserving logic. The study also revealed basic arithmetic lives in the memorization pathways.
#MLSky
#MLSky
Researchers isolate memorization from problem-solving in AI neural networks
Basic arithmetic ability lives in the memorization pathways, not logic circuits.
arstechnica.com
November 12, 2025 at 4:15 PM
Goodfire.ai researchers isolated memorization from reasoning in neural networks. They found they could surgically remove verbatim recall while preserving logic. The study also revealed basic arithmetic lives in the memorization pathways.
#MLSky
#MLSky
You won’t believe the excuses lawyers have after getting busted for using AI
arstechnica.com/tech-policy/...
#LawSky
arstechnica.com/tech-policy/...
#LawSky
You won’t believe the excuses lawyers have after getting busted for using AI
I got hacked; I lost my login; it was a rough draft; toggling windows is hard.
arstechnica.com
November 12, 2025 at 4:08 PM
You won’t believe the excuses lawyers have after getting busted for using AI
arstechnica.com/tech-policy/...
#LawSky
arstechnica.com/tech-policy/...
#LawSky
AI and quantum computing are reshaping cybersecurity. AI enables faster, large-scale attacks, while quantum computing threatens to break current encryption. Proactive defense requires a zero trust approach, using AI to fight AI and preparing for post-quantum cryptography.
#MLSky
#MLSky
Reimagining cybersecurity in the era of AI and quantum
The threat landscape is being shaped by two seismic forces. To future-proof their organizations, security leaders must take a proactive stance with a zero trust approach.
www.technologyreview.com
November 12, 2025 at 3:57 PM
AI and quantum computing are reshaping cybersecurity. AI enables faster, large-scale attacks, while quantum computing threatens to break current encryption. Proactive defense requires a zero trust approach, using AI to fight AI and preparing for post-quantum cryptography.
#MLSky
#MLSky
🧪 'Godfather of AI’ becomes first person to hit one million citations
www.nature.com/articles/d41...
#AcademicSky
www.nature.com/articles/d41...
#AcademicSky
‘Godfather of AI’ becomes first person to hit one million citations
The milestone makes machine-learning trailblazer Yoshua Bengio the most cited researcher on Google Scholar.
www.nature.com
November 12, 2025 at 3:40 PM
🧪 'Godfather of AI’ becomes first person to hit one million citations
www.nature.com/articles/d41...
#AcademicSky
www.nature.com/articles/d41...
#AcademicSky
Should ageing be classified as a disease? Some researchers say yes, arguing it would unlock vital funding and speed up drug approval. Others warn of ethical risks, potential for increased ageism, and distracting from lifestyle interventions.
#MedSky 🛟
#MedSky 🛟
Is ageing a disease? The debate that could reshape medicine
Classifying ageing as a disease might bring funding and other benefits, say some researchers, but others say this is fraught with ethical and regulatory implications.
www.nature.com
November 12, 2025 at 3:23 PM
Should ageing be classified as a disease? Some researchers say yes, arguing it would unlock vital funding and speed up drug approval. Others warn of ethical risks, potential for increased ageism, and distracting from lifestyle interventions.
#MedSky 🛟
#MedSky 🛟
Analysis of Fox News coverage preceding the Portland troop deployment reveals a misleading narrative.
ProPublica found the network used 2020 footage labeled as 2025 and described protests as "riots" contrary to official police designations.
ProPublica found the network used 2020 footage labeled as 2025 and described protests as "riots" contrary to official police designations.
“Riots Raging”: The Misleading Story Fox News Told About Portland Before Trump Sent Troops
After reviewing coverage from the network and hours of social media videos that preceded Trump’s decision, ProPublica found that Fox’s portrayal of “Portland rioters” routinely instigating violence wa...
www.propublica.org
November 12, 2025 at 2:14 PM
Analysis of Fox News coverage preceding the Portland troop deployment reveals a misleading narrative.
ProPublica found the network used 2020 footage labeled as 2025 and described protests as "riots" contrary to official police designations.
ProPublica found the network used 2020 footage labeled as 2025 and described protests as "riots" contrary to official police designations.
The city glow below, the cosmos above.
Worth staying up late for this spectacular display of the #NorthernLights
Worth staying up late for this spectacular display of the #NorthernLights
November 12, 2025 at 2:40 AM
The city glow below, the cosmos above.
Worth staying up late for this spectacular display of the #NorthernLights
Worth staying up late for this spectacular display of the #NorthernLights
November 12, 2025 at 2:06 AM
We are working on a new paper, and one of the lead authors wanted to try for The Lancet.
We took at look at the fees, and.. yikes!
We are no longer considering Lancet.
We took at look at the fees, and.. yikes!
We are no longer considering Lancet.
November 11, 2025 at 9:29 PM
We are working on a new paper, and one of the lead authors wanted to try for The Lancet.
We took at look at the fees, and.. yikes!
We are no longer considering Lancet.
We took at look at the fees, and.. yikes!
We are no longer considering Lancet.
A survey at a large medical school reveals a gap in AI adoption. Most faculty and students identify as novice users, citing barriers like lack of knowledge and unclear benefits. Both groups are calling for structured training and ethical guidance.
#MedSky #MLSky
#MedSky #MLSky
Integrating artificial intelligence into medical education: a roadmap informed by a survey of faculty and students
We surveyed faculty and students at a large urban medical school to assess their awareness, usage patterns, and perceived barriers to AI adoption, aiming to identify opportunities for meaningful…
pmc.ncbi.nlm.nih.gov
November 11, 2025 at 5:07 PM
The FDA Digital Health Advisory Committee seemed stunned by agency questions on regulating GenAI for mental health.
The panel expressed deep reservations about AI chatbots treating MDD, particularly hypothetical OTC versions or use in children.
#MedSky #MLSky
The panel expressed deep reservations about AI chatbots treating MDD, particularly hypothetical OTC versions or use in children.
#MedSky #MLSky
FDA questions on genAI-enabled chatbots raise concerns from expert panel
During the second meeting of the US Food and Drug Administration@s (FDA) Digital Health Advisory Committee (DHAC), members of the panel seemed shocked and, at times, lost for words when grappling with the agency@s discussion questions surrounding the potential regulation of generative artificial intelligence (AI)-enabled digital mental health medical.
www.raps.org
November 11, 2025 at 4:54 PM
I installed Govee lights on our house, and being able to simply turn them on during the holidays has been amazing!
The Best Permanent Outdoor Lights (2025): Govee, Eufy, Lepro
The Best Permanent Outdoor Lights (2025): Govee, Eufy, Lepro
We Found the Permanent Outdoor Lights Worth Putting on Your Home
DIY permanent eave lights that attach to the exterior of your home are becoming more popular. But are they worth the effort (and cost)?
www.wired.com
November 11, 2025 at 4:01 PM
I installed Govee lights on our house, and being able to simply turn them on during the holidays has been amazing!
The Best Permanent Outdoor Lights (2025): Govee, Eufy, Lepro
The Best Permanent Outdoor Lights (2025): Govee, Eufy, Lepro
Our @lastweektonight.com “Cabernet Savignjohn” has arrived!
Sales sent $50k support 16 California food banks!
Sales sent $50k support 16 California food banks!
November 11, 2025 at 4:06 AM
Our @lastweektonight.com “Cabernet Savignjohn” has arrived!
Sales sent $50k support 16 California food banks!
Sales sent $50k support 16 California food banks!
Finally can take off the mask and return back to normal!
November 10, 2025 at 5:49 PM
Finally can take off the mask and return back to normal!
From @benjedwards.com: new research reveals AI struggles to mimic humans on social media. A "computational Turing test" achieved 70-80% accuracy. The persistent tell is not intelligence, but emotion: AIs are too polite and fail to replicate human negativity. #MLSky
Researchers surprised that with AI, toxicity is harder to fake than intelligence
New “computational Turing test” reportedly catches AI pretending to be human with 80% accuracy.
arstechnica.com
November 10, 2025 at 5:19 PM
From @benjedwards.com: new research reveals AI struggles to mimic humans on social media. A "computational Turing test" achieved 70-80% accuracy. The persistent tell is not intelligence, but emotion: AIs are too polite and fail to replicate human negativity. #MLSky
AI leaders predict rapid, powerful advances. A new expert panel (LEAP) offers specific, testable forecasts. The panel is skeptical of near-term AGI but anticipates massive long-term impact. Yet, recent benchmarks suggest even experts underestimate AI's pace.
#MLSky
#MLSky
A new project aims to predict how quickly AI will progress
Superforecasters weigh in on the subject
www.economist.com
November 10, 2025 at 5:07 PM
AI leaders predict rapid, powerful advances. A new expert panel (LEAP) offers specific, testable forecasts. The panel is skeptical of near-term AGI but anticipates massive long-term impact. Yet, recent benchmarks suggest even experts underestimate AI's pace.
#MLSky
#MLSky
Seven more families are now suing OpenAI over ChatGPT’s role in suicides, delusions
techcrunch.com/2025/11/07/s...
techcrunch.com/2025/11/07/s...
Seven more families are now suing OpenAI over ChatGPT's role in suicides, delusions | TechCrunch
In one case, 23-year-old Zane Shamblin had a conversation with ChatGPT that lasted more than four hours.
techcrunch.com
November 10, 2025 at 5:05 PM
Seven more families are now suing OpenAI over ChatGPT’s role in suicides, delusions
techcrunch.com/2025/11/07/s...
techcrunch.com/2025/11/07/s...
Reports from the EPA describe internal chaos amid the shutdown. Furloughs appear haphazard, with little information provided to staff. Meanwhile, work on specific deregulatory priorities continues, even as other agency functions cease.
#MedSky 🛟
#MedSky 🛟
The EPA Is in Chaos
“We learn who is furloughed when we send an email to someone and get the out-of-office message,” one employee tells WIRED.
www.wired.com
November 10, 2025 at 4:59 PM
Reports from the EPA describe internal chaos amid the shutdown. Furloughs appear haphazard, with little information provided to staff. Meanwhile, work on specific deregulatory priorities continues, even as other agency functions cease.
#MedSky 🛟
#MedSky 🛟