Where new things come from. Below, everything that touches this field, including pieces that live mostly in neighboring territory.
With all the healthy chatter about "verifiability" ignited by Andrej Karpathy, I want to point out that even if "verifying" (a program) can be hard, it is still easier than coming up with a good idea for a…
I have been thinking about dating for 25 years. As in, how does someone find the right partner for their life? Incidentally, I found mine on a dating app AFTER meeting her in real life, which made for an…
Alexander Fleming's discovery of penicilin is a classic example of serendipitous discovery. "Fleming kept a messy lab. He left petri dishes, microbes and nearly everything else higgledy-piggledy on his lab…
Alexander Fleming's discovery of penicilin is a classic example of serendipitous discovery. "Fleming kept a messy lab. He left petri dishes, microbes and nearly everything else higgledy-piggledy on his lab…
Serendipity has been, and still is, a most intriguing concept to me, central to my philosophy of life as well as to pretty much all my research -both as a topic AND as an engine of progress. The core principle…
As a serendipity engineer, I am always looking for examples of discoveries made serendipitously. The quintessential example is Alexander Fleming's stumbling upon penicillin, a discovery of monumental import to…
In a recent perspective in Nature, Thomas Fink of the London Institute for Mathematical Sciences makes an interesting case for why the creativity of (good or great) mathematicians is, and will remain for some…
Allie K. Miller is pointing to the trigger point of a Cambrian explosion in innovation. People (myself included) have been criticizing the "chatbot interface" but it turns out that it is a super effective way…
The conviction that serendipity can be, if not systematized, at least harnessed for augmented scientific discovery, has fueled my (obviously nonlinear) professional trajectory. As we saw with Fleming's…
Understanding Cognitive Biases Through a Sampling Limitations Framework: A Comprehensive Review and Theoretical Integration Abstract This paper presents a comprehensive theoretical framework that…
Eli Lilly's Chorus unit, established in 2002, emerged as one of the most successful innovations in pharmaceutical R&D productivity, achieving 3-10x productivity improvements over traditional development models…
The original “fitness beats truth” (FBT) theorems do not require an explicit assumption that “seeking truth is costly.” The key requirement, instead, is that evolution favors any perceptual or cognitive…
Recall and recognition are two ways of retrieving the same stored information, and they differ sharply in how much work they demand. Recall means producing a target from memory with little or no external cue.…
Based on the handwritten notes in the image, the text is organized into a large mind map titled "The Future Belongs to Curators". Here is a detailed extraction of all the notes, grouped by their apparent…
That's the title of a short IEEE Spectrum piece ( and a case in point as a second follow-up to Hugging Face's Thomas Wolf's post on how mediocre current AI systems, particularly LLMs, are at asking…
“More people: 100x more people will gain the confidence to express their ideas visually and engage with creative tools again (most of whom abandoned drawing as children and never attempted creativity again).…
A great take on genAI hallucinations by Colin F. "Hallucinations", however we define the concept, are a core property (and perhaps the core value proposition) of LLMs and genAI systems. My view: Trying to root…
I saw a post by @juergen schmidhuber recently ( that gave me pause. About convolutional neural networks and the neocognitron, a 1979-80 neural network architecture by the Japanese scientist Kunihiko Fukushima…
In his 60 minutes interview with Anderson Cooper two years ago, legendary music producer Rick Rubin gave viewers, as usual, incredible insights into his success and also about the future (of humans and AI).…
With all the healthy chatter about "verifiability" ignited by Andrej Karpathy, I want to point out that even if "verifying" (a program) can be hard, it is still easier than coming up with a good idea for a…
Raffi Krikorian this is great! Gen AI can "generate" (an infinite amount of) novel things riffing on a theme, but I am sure that humans were involved in curating what was carved into stone. That's augmented…
? I periodically remind myself to re-read this 2012 by Cormac Herley at Microsoft, or at least the abstract. The answer to the question, in a nutshell, is that if someone responds, they are much more likely to…
I don't know Lauren Ducrey but she recently posted what I think is the most concise summary of the uses of genAI: "This is my biggest question when I'm using AI: Am I being lazy or am I being curious?" (Lauren…
You can always count on Derek Lowe for an incisive (meaning, with tooth) analysis of a half-baked paper with a promising title ( While Derek focuses on the paper's poorly defined categories (can a target…
In a conversation with economist Tyler Cowen last month, Peter Thiel expresses his view that LLMs will make things "worse for the math people than the word people." He continues: "What people have told me is…
"But, in truth, the David who solved the K(π,1) conjecture is a social parasite of the much better mathematician, the David who crafted Definitions 2.4 and 9.3." A post by David Bessis (by way of Alexander…
Thank you Kes Sampanthar for your read on this paper. All the assumptions made by Daron, all the things he neglects, suggest a linear approximation whereby he applies a very short term gradient (most gains…
Vemödalen The Fear That Originality Is No Longer Possible Day 2 in John Koenig's rabbit hole, The Dictionary of Obscure Sorrows, another concept that resonates in the age of AI everywhere. Now the PhD…
One irregular dispatch when a new piece lands: notes on AI, evolution, complexity, and the biology of discovery. No noise.
Delivered with Buttondown · unsubscribe anytime