How AI might slow scientific progress

/

“One of my growing concerns is that A.I. could inadvertently slow scientific progress. The theoretical physicist Max Planck is often credited with saying that “science advances one funeral at a time.” I am mindful that I may be quite wrong in my viewpoints. However, if my opinion becomes encoded into A.I. systems and persists indefinitely, will it hinder the evolution of new scientific ideas?” - Tamara Kolda, who runs MathSci.ai, a consultancy in the San Francisco Bay Area, quoted in the New York Times

AI Definitions: Imitation Learning

/

Imitation Learning – This is a popular method for training robots, along with reinforced learning. The robots learn by watching humans or by being given data on other robots which are being operated by humans. Out of fashion for decades, it has recently come back into favor in robotics because of AI. The downside to this technique is the need for large amounts of data for the robots to imitate new behaviors.

More AI definitions

18 Recent Articles about the Impact of AI on Health Care

/

Health Advice From A.I. Chatbots Is Frequently Wrong, Study Shows – New York Times

As AI enters the operating room, reports arise of botched surgeries and misidentified body parts – Reuters

ChatGPT can analyze Apple Watch health data. Here’s how a doctor views it. - The Washington Post

Why some hospitals are making their own ChatGPTs for patient records – Stat News

A.I. Is Making Doctors Answer a Question: What Are They Really Good For? – New York Times

‘Dangerous and alarming’: Google removes some of its AI summaries after users’ health put at risk – The Guardian

I let ChatGPT analyze a decade of my Apple Watch data. Then I called my doctor. – MSN

Institutions are missing AI’s real potential for drug discovery – Semafor

AI-generated sensors open new paths for early cancer detection - MIT News

Your next primary care doctor could be online only, accessed through an AI tool – NPR

What are the limits to biomedical research acceleration through general-purpose AI? – Nature

OpenAI launches ChatGPT Health, encouraging users to connect their medical records – The Verge

Utah permits nation's first AI drug prescriptions – Axios

Hospitals Are a Proving Ground for What AI Can Do, and What It Can’t – Wall Street Journal

Where Is All the A.I.-Driven Scientific Progress? - New York Times

40 million people turn to ChatGPT for health care - Axios

Researchers create a machine learning model “to distinguish paper mill publications from genuine cancer research articles.”- The BMJ

In China, A.I. Is Finding Deadly Tumors That Doctors Might Miss - New York Times

AI Definitions: OpenClaw AI

/

OpenClaw AI – This open-source artificial-intelligence agent is designed to assist users with everyday tasks, such as reading and sorting email, scheduling calendar events, and making purchases. Text it on a variety of platforms and it will remember your messages and preferences, send you reminders and automate tasks for you. While its creator promises a capable assistant, critics warn it is not a polished, enterprise-ready product but a rough outline of a tool with significant security concerns, especially since it is still at work after users have logged off. OpenClaw was released as open-source software on the platform GitHub in November of 2025 under the name Clawdbot, which was briefly changed to Moltbot.

More AI definitions

AI Definitions: GEO (Generative Engine Optimization)

/

GEO (Generative Engine Optimization) – Also known as “answer engine optimization” (AEO), this is the process of optimizing content to boost its visibility in AI-driven search engines (ChatGPT, Perplexity, Gemini, Copilot and Google AI). While SEO helps brands increase visibility on search engines (Google, Microsoft Bing), GEO is all about how brands appear on AI-driven platforms. There is overlap between the goals of GEO and traditional SEO. Both use keywords and prioritize engaging content as well as conversational queries and contextual phrasing. Both consider how fast a website loads, mobile friendliness, and prefer technically sound website. However, while SEO is concerned with metatags and links in response to user queries, GEO is about quick, direct responses from synthesizes content out of multiple sources.

More AI definitions

He led safeguarding at a major AI company. His next career move will surprise you.

/

Mrinank Sharma, who had led Anthropic’s safeguards research team since its launch last year, shared his resignation letter in a post on X Monday morning, which quickly garnered attention and has been viewed 1 million times. In his letter, Sharma said the “world is in peril,” not just from AI, but a “whole series of interconnected crises unfolding in this very moment.” After leaving Anthropic, Sharma said he may pursue a poetry degree and “devote myself to the practice of courageous speech,” adding he wants to “contribute in a way that feels fully in my integrity.”

Read more at Forbes

16 Articles about Social Media & AI

/

AI Definitions: Convolutional neural networks 

/

Convolutional neural networks (CNNs or ConvNet) – These deep learning artificial neural networks, often used in computer vision for object recognition, are trained on thousands of images. It works similarly to how our human eye processes images. The network is trained to recognize "kernels," which are tiny pieces of an image. However, they can fail when they encounter the same objects under new lighting conditions or from a different angle. CNNs play a role in unlocking our phones with our faces, identifying road signs in self-driving cars, and automatically tagging people in our photo galleries. CNNs were first introduced in 1989 by NYU professor Yann LeCun and have been used with autonomous vehicles and security camera systems.

More AI definitions

16 Articles about Students Using & Impacted by AI

/

AI Definitions: Agentic AI

/

Agentic AI – Able to operate more independently than AI Agents, Agentic AI operates like a workflow, able to adjust its strategy and continuously learn as it encounters different situations. Agentic AI systems aren't passive tools waiting for input or mere automation. They can update plans based on intermediate findings without needing continuous human supervision. It’s not just following the rules as agents do, Agentic AI is supposed to be a colleague that can analyze a problem, propose a plan, and take action. Think of agentic AI as a team of digital colleagues where some agents are coordinators and some are specialists. It might call out to additional models or external systems, such as a search engine or querying a database to complete a task. This can be particularly effective in data-heavy fields such as biology, chemistry, and drug discovery. On a personal level, instead of simply helping you find a hotel room to book, agentic AI can plan the trip if it is given access to programs with your schedule and preferences. Agents can better handle the back-and-forth interactions that most real workflows require than rule-based systems. Despite its capabilities, AI agents can struggle in open-ended or unpredictable environments, especially when tasks lack clear structure or context. It will likely take years to for most agentic AI systems to be tailored to specific industries or problems.

More AI definitions

How Many Friends Can We Have?

/

The number of people with whom we can maintain a stable relationship is about 150, according to British anthropologist Robin Dunbar.

He says: “We devote around 40 percent of our available social time to our 5 most intimate friends and relations … and the remaining 60 percent in progressively decreasing amounts goes to the other 145.”

Friendship is the single most important factor influencing our health, well-being, and happiness. Creating and maintaining friendships is, however, extremely costly, in terms of both the time that has to be invested and the cognitive mechanisms that underpin them. Part of friendship is the act of mentalizing, or mentally envisioning the landscape of another's mind. Cognitively, this process is extraordinarily taxing, and as such, intimate conversations seem to be capped at about four people before they break down and form smaller conversational groups.  

Read more at the BigThink

9 Things People are Trying to Get AI to Do

/

Rent a Human

/

Rent a Human.ai is a new site promoted as a place to book humans for real-world tasks your AI can’t do. “According to the site, more than 81,000 "rentable humans" have already signed up to offer paid services to bots. The tasks themselves range from mundane errands like picking up packages to holding signs or delivering flowers to Anthropic. Rent-a-Human requires users to connect crypto wallets in order to get paid.” More at Mashable

AI Definitions: GPT

/

GPT (Generative Pre-trained Transformer) – GPT refers to a LLM (large language model) type of AI that first goes through an unsupervised period (no data labeling by humans) followed by a supervised "fine-tuning" phase (some labeling). G is for Generative because it generates words. P is for Pre-trained because it’s trained on a lot of text. This step is called pre-training because many language models (like the one behind ChatGPT) go through important additional stages of training known as fine-tuning to make them less toxic and easier to interact with. T stands for Transformer which is a relatively recent breakthrough in how neural networks are wired. They were introduced in a 2017 paper by Google researchers, and are used in many of the latest AI advancements, from text generation to image creation.

More AI definitions