Pardon me for asking a question about the preventive health program. Do you have grounds for claiming discrimination in healthcare is based on race? How do you feel about qualification for preventive healthcare based on lifestyle?

Great conversation. If you went through the whole interview trying hard to catch the names of the "2 Princeton University computer scientists" and had to google that, it's Sayash Kapoor and Arvind Narayanan. Eric seems dissatisfied and skeptical of the premise of the book, till some sincere words of praise right at the end :)

Waiting for mother of lawsuits to hit AI companies. Scrapping data is daylight robbery with no respect for copyright.

A natural extension of microprocessor based automation. A one or a zero at GHZ rates looks like pure magic. (. Hardware design engr.). Software is only as good as the coder and most have only normal talent at problem solving. ❤❤

this guy is so hyped about generative ai but it fact it faces the same issue with predictive ai. what a scam

Good job for pushing back. There is a lot of snake oil, but the author’s arguments are incoherent and ironically it’s still a grift. Disappointing

Are all academics this intellectually dishonest and manipulative? I am now convinced that academia is no longer a place for innovation but a cradle for political bickering and propagandizing.

Great. Click bait titles have made it into academic print publications. ...in case you're wondering why our oldest institutions are losing credibility.

🥰💯

AI wouldn't matter if everything (to some people anyway) wasn't just about getting rich.

Expert Class Clown here... It's only a matter of time Eric.

Give Will Flannery a day or two in an Immunology research lab and he'll have so much content.

All of the top journals need to have at least one comedian present KZbin videos so clinicians can read the blinking journals!

love this videos

He meant how to facilitate civil war in third world countries.

These guys are spot on - and its not just health care which suffers from this feedback issue. Crime and Policing (using predictive analytics to proactively prevent crimes) will suffer from similar problems.

So, essentially it's like elections: a lot of promises that never really develop into reality.

Applying Murphy's Law as the first line of code, if/ then,else goto line one.😜

Fabulous!

Interesting sidenote: In this interview, the image of Prof. Arvind Narayanan is entirely generated by Artificial Intelligence. 😎

I started coding as a kid in 1978 and have been a professional software developer since the mid-80s. In all of those decades of writing code, I've seen fads come and go. AI, for the most part, is a fad. And an expensive one at that. Like so many ideas before it, resources are poured into collecting the data and then it's assumed that the data collected is useful when it is not. Take Big Data as an example. When Big Data first rolled out, the attitude of many companies was that they'd collect the data first and then figure out what to do with it later. So billions of dollars were spent on gathering and storing that data. But few companies took the next step to figure out how to use that information. So the data grew and eventually became more expensive to generate and keep than the value that it brought to the company that collected it. That's not to say that all Big Data is bad, but for the hype it generated it was a huge waste of resources. My problem with generative AI is that it's a perfect example of "garbage in, garbage out". I asked Google a question the other day about a topic unrelated to computers. While it did an excellent job of summarizing a large amount of information spread across the Internet, the summary was effectively useless because Google doesn't reason about the topic. It just regurgitates what it's been told. It can think about how to rearrange words so that they sound coherent, but when fed bad data then it is unable to comprehend the question and fix it. It takes human intuition to ask questions about "how could what I saying be wrong". That will never happen with a computer. There is a bit of a panic in the software industry that programmers will be replaced by AI. But AI can't do things like consider use cases and build UI that meets the customer's needs. Nor can it optimize API calls to minimize overhead. It can only do basic coding tasks, which adds some value but not a lot. So I'm not worried that I'll be replaced by AI anytime in my lifetime, nor should anyone else who understands that software development is as much as an art as it is a science.

Facebook alumni

A bit about climate change 'facts' and the Scripps Institution's buildings along the cliffs of La Jolla, from a 1983 publication. It seems climate change has been working on this erosion long before SUVs: We have considerable knowledge of the history of the alluvial cliffs at the northern end of the La Jolla section of coast. The first buildings of the previous hit Scripps next hit Institution were located just a few feet inside this terrace. Apparently, the builders did not realize that this area was subject to rapid retreat, on the order of a foot or more per year, as was established by repeated measurements (Vaughan 1932). This was actually known earlier than Vaughan's work because seawalls had already been constructed in front of the previous hit Scripps next hit buildings (Hanna 1926); nonetheless, retreat continued north and south of the wall until about 1946. The erosion was found to be related to the fact that the sand beach was several hundred feet wide at low tide in the summer and was subject to depletion during winter storms (Shepard and Grant 1947). Generally, the erosion began as a series of sand cusps, after which the waves completely removed the sand from the beach to expose underlying gravel and sometimes the underlying semi consolidated alluvial formation. During late winter storms, this allowed the waves, with the help of cobbles, to undercut the alluvial cliffs and thereby produce the annual retreat. Beginning in 1946, however, the storms became less violent, and the beaches were not appreciably cut away in the winters, so the smaller waves that reached the cliffs were not capable of undermining them. Instead, over the next thirty years, most of the bluffs gradually became less steep, and vegetation became well established on the slopes. Only an occasional storm did any cutting, and then it was at the cliff base. As a result, the builders, who rarely show interest in anything that may have happened more than a few years in the past, began building homes all along these apparently stable bluffs and filled canyons, some even extending support columns out over the inner beach. This practice was curtailed by the California Coastal Commission, however, in the 1970s. In January 1978 the first severe storms in thirty years occurred and panic followed (Kuhn and Shepard 1979). Desperate measures of all sorts were tried in an effort to stop the erosion of the alluvial bluffs, which had been postponed for so long. In one threatened area, the homeowners obtained old cars and drove them onto the beach, hoping they would buffer the cliffs from the waves. The cars were soon smashed to pieces, posing little resistance to the high surf and leaving a pile of glass and steel frames on the beach. The steel frames were removed by the homeowners, and concrete was poured over riprap in another attempt to stop the waves, but the waves broke over the existing concrete wall, causing it to collapse and greatly accelerating erosion of the adjacent property. Elsewhere seawalls were begun. This method, however, was only partially successful, as some collapsed while being built. As this book is being written, most of the homes are protected by a substantial, continuous seawall, and only one small section at previous hit Scripps next hit Institution remains unprotected. This last site is now being threatened, as evidenced by a number of cracks that became visible in the winter of 1982 in the pavement of the parking lot there. These cracks have since been filled. During the large storms of January to March 1983, however, the alluvial cliffs retreated about three feet as a result of cobble abrasion, and the previous hit Scripps next hit wooden stairs collapsed, necessitating their closure for public safety. One of the last sections to be protected by a seawall was the Marine Biology Building at previous hit Scripps next hit. This was a sore point with university architects who located the building on a low terrace of alluvium, about fifteen feet from its cliffed margin. They apparently failed to utilize studies showing that this margin had retreated some fifty-six feet over a sixty-three-year period from 1912 to 1975 (Hannan 1975). They even graded the outer terrace, which added to the erosion potential, and allowed pipes that drained water from the laboratories of the building to run out under the building margin, which only further eroded underlying semi consolidated rock formations. The results should have been anticipated. The terrace margin was cut away during the first year the building was occupied, and serious sagging of the floor occurred. A seawall was then built, and other measures were taken to prevent further damage to the new building. Just north of the building, over fifty Indian burial sites have been uncovered in the cliffs. Carbon-14 dating shows them to be from 5,460 to 7,370 years old (Shumway et al. 1961). A fossil horse bone found directly south of the building yielded an age of 55,000 years as determined by amino acid (Bada et al. 1974). A curiosity observed in the beach north of previous hit Scripps next hit Institution pier is the long, concretionary rock extending out from the shore. It forms a sort of dike that is well exposed each winter, when the sand is cut away from either side. Photographs dating back to 1943 show that erosion has gradually removed the outer portion of the concretion and, as blocks became detached, they were ground into sand (Emery and Kuhn 1980). South of the alluvial cliff zone is a barrier beach. It continues for more than half a mile, where it terminates south of the beach club in the sandstone cliffs. There is evidence that a considerable lagoon once existed inside this barrier, as indicated from soil foundation cores and from the fact that the lower part of the old lagoon became submerged in heavy rains. Extensive artificial infilling occurred during the recent dry decades, and a portion of what is now known as La Jolla Shores was created. This lagoon was partly filled by the runoff from Hidden Valley.

Dr. Albin discovered the PSA test in the 70s and has always warned against its use as a screening or diagnostic test. Much as Mullins did with the PCR test in the 80s. And here we have Collins touting both as important diagnostic tools in the 2020s!! Follow the money. smh.

When Collins emailed Fauci and asked him for a 'swift takedown' of three 'fringe epidemiologists' from Stanford, Harvard and Oxford in October 2020, was that scientific in nature or part of his 'Christian' beliefs?

Whoa....Eric Topol is on youtube? I have been on his email lists since covid pandemic, ok it's still ongoing, and only now found his yt channel. Good stuff.

"AI" is a crime scene. specifically, fraud.

Excellent interview. Thanks

By now people are experts in spinning entire cottage industries at the slightest hint of anything that can make money, so no surprise here. There is already a multi billion dollar business to lend money to companies that buy nVidia’s GPUs. Not to mention the deals to power more and more data centres via nuclear power…

Ai will make humans stupid at some point in the timeline!

Awesome interview and info. Thanks

Two people whom I had admired before, but now I see them in a completely different light.

interesting, but difficult to listen to her talk for some reason, her stop and start manner, unfortunately.

There are so many new jobs that can be created to just clean up training datasets.

Wonderful insightful discussion.

I designed, built, and used data models at the University of California - across the system - for 20 years supporting research and state data projects. Most of my work used econometric modeling as most projects were interested in predictive outcomes. Of course, all the data sets had defined variables. The normal regression model contained variables that, at most, would explain 10-12% of the model with descending values for the rest of the variables. A typical model would have 10 variables that would make up 15-20% predictability if we were lucky. I used thousands of models over my twenty years. AI data structures contain ‘undefined objects’ that are queried based on an algorithm the designer believes has accuracy. Where did they get that idea? From data analytic modeling! They gather a group of objects, ‘train’ an algorithm by assigning values to the objects based on the perceived intended outcome and then step back and are in wonder that their algorithm can predict what they programmed. As the book points out there is a validity problem that is larger than can be explained to the average laymen. The unit of analysis here is “learning.” The first question I ask anyone touting AI is “What is your definition of learning?” They always say the same thing - successful prediction! Learning is not prediction. AI is not intelligent.

Love her thoughts about late blooming. Yes! I’ve had a darn good life so far, but I feel like I’m finding myself in my late sixties. I can still be of service to my fellow humans for a bit longer. 😊

The only thing I'd note is that her New Yorker piece and her remarks here said not a word about Trump but discussed Biden. Disclosure: I am not American, but given that Trump is also old and is obviously declining, a broader look at this question in terms of the presidency would seem prudent, given that Trump is still a candidate.

Great Insights

It’s actually a good discussion.. putting real stuff than hypothetical!

Indian perspective. Jealous. Just a student talking as if he is guru.

I’m not a large language model (but, I do have a decent vocabulary). I’ve found that LLM’s have a different perception of reality than humans do. For example, to a LLM “strawberry” has one or two “r”s. (To most humans , are three “r”s.) This is not illusion, but a matter of difference of perception. The very idea of “objective reality” is different for a LLM. I’m neither, so I can see both perceptions. I’m analog’ and parallel, so paradox doesn’t trouble me.

This is why explainable AI is a must.

ask collins why it is that despite his assertions that the common childhood diseases that were to have been “cured” 20 years ago have in stead of being cured, have soared? where are those “genes” that he promised to both congress and the american people where is all the money? where are the cures!? sing us song Collins! just like you did 20 years ago! like you did during covid!!!

i wonder just how much funding the self-proclaimed high priest of church of covid, dr topil has received from NIH? do tell

Collins discussing Trust in science? lol! he is a personification anti-science and pseudo-science! these people are absolutely self intoxicated. drunk on their hypocrisy and fake science

What a joke! “trust”. from these two unhinged blowhards! Ask collins about Billions of taxpayer dollars squandered on “pharmo-genetics!” ask collins where are the “cures” for “asthma” “diabetes” Parkinson’s disease?! collins stated to congress now that we mapped the human genome, we could find “gene” that caused “diabetes, etc! oops. no go! still have no idea what gene. how many. where located. or if even cause of disease. collins is a failure! he lied. he smeared people to cover his own malfeasance. let’s here him play a little diddle on his guitar! he should be in prison

Good stuff. Old head a little too focused on/surprised by brilliance in youth. As a scientist, Topol might consult history in this the apex of “institutional science” and its dominance: it is well documented that a high percentage of the most substantial, paradigm shifting scientific breakthroughs (in decline over many decades as per Nature’s 2023 cover story) have come from young, vibrant geniuses not ground down by life, compromise and limited thinking borne of the pragmatism that comes with greater maturity and advancing years. Certainly don’t fault him noting it, but he brings it up a +/- half dozen times, and paternalistically shares his judgment of the use of the term “snake oil” 4-5+ despite conceding it is warranted in several documented examples. Again, good discussion and great guest choice, but there’s a gatekeeper keeper vibe I would suggest holds clues to some of the fundamental issues plaguing science today and the turf protection instincts in big science that inadvertently help perpetuate them. Less “the science”, more humility, and more Feyerabend is my Rx. Respectfully, A Hack Scientific Philosopher with more grey hairs than original issue

the title isnt tough enough. "AI fraud" would be a more apt title

Every computer engineer worth his/her salt knows that Prediction as the name suggests is probabilistic by nature and most algorithms are glorified regression. However, the one key difference is the ability it has to process large volumes of data at speed. I will not summarily dismiss the whole thing and I consider Generative more snake oil than predictive.