Getting clocks right - AI almost there

 kw: experiments, art generation, artificial intelligence, simulated intelligence, clock dials

It takes a child about six years to learn to read an analog clock. It takes even longer to learn to draw one correctly. Art generation software seems to be on the verge of reaching this milestone.

This is the winning image in the experiment I describe below.

The prompt read

A beige wall in a store selling clocks, showing four circular wall clocks, each reading exactly 3:27.

There are two parts to showing the correct time on an analog clock dial. Firstly, the minute hand has to point to the correct minute mark. Here, all four clock dials show this correctly.

Secondly, the hour hand must point to the correct location. In this case, the hour hands should all point 27/60ths (45%) of the distance between the 3 and the 4. I made careful measurements on the original image. The four hour hands point identically to a spot just above the second of the marks between 3 and 4 (~38%). They should point just a little below it. They all are where they should be at 3:23, about six minutes too early. There is one other possible anomaly in the image. The fifth clock shown, near lower left, reads 10:10, as nearly all advertisements for clocks have for decades. This abundance of training material causes clocks reading 10:10 (or sometimes 1:50) to dominate images of clocks produced by nearly all art generators.

This image is the best result that an art generation model has achieved since I began using DALL-E2 almost four years ago. This model is named GPT Image 2. I believe it is the model used when you ask ChatGPT 5.4 to create an image. At this point, GPT earns an A- from me.

For this experiment I tested ten art engines, GPT Image 2 and nine others. I had each of them produce two images and selected the best one to be shown below. My focus was the most recent models offered by Leonardo AI and OpenArt. There is a lot of overlap between these two "umbrella" sites. I began with Leonardo AI, using GPT Image 2 first (The other image by GPT was nearly this good, but the four clocks all pointed at 3:25, and all four hour hands were pointed the same as these). Now I'll discuss the other models tested, in order.

Google's flagship model is Nano Banana 2. It excels in realistic imagery. In this case, it set the beige wall I prompted as though it were above a pass-through into the workshop of the clock store, nicely blurred for a bokeh effect.

The times are not all the same, and none of them is 3:27. They range, according to the minute hands, from 3:09 to 3:18. None of the hour hands is correctly aimed, but a couple of them are close.

I had noticed earlier than when I prompt Nano Banana 2 for an image including a clock, and specify a time to show, the result is only approximately correct. However, I've usually accepted that because for the scene I am creating, almost any time other than 10:10 will do. So for other reasons, I use Nano Banana 2 at least as frequently as GPT Image 2.

The Ideogram model is a private brand. It was among the first to be able to correctly render text (most of the time). This version is P-Image Ideogram.

Text: good if I had asked for any, based on prior experience. Clock dials: not so good. All show 10:10, and even more so, the hour hands point almost exactly at the 10, where they should be one-sixth of the way from the 10 to the 11.

The dial designs are interesting. The blue and black ones are quirky, mixing genres in a way few clock designers would do, and the blue one in particular has inaccurate Roman numerals.

The Seedream series is owned by ByteDance; this is version 5.0.

The minute hands are exactly right. The hour hands have all gotten ahead of themselves, pointing almost all the way to the 4, as though the time were 3:57 rather than 3:27.

Although the other clocks shown in the background are bokeh-blurred, I can see that each reads a different time. This is a plus for Seedream, something I can take advantage of when I actually need multiple clocks in an image, such as depicting a clock repair business.

Finally, under the Leonardo AI umbrella, Lucid Origin is a Leonardo "house brand." It is perhaps a little dated, but new enough to be quite good with many imaging tasks.

As far as showing the time, although these clocks don't show 10:10, the model has entirely gone off the rails. None of the four pseudo-times shown is close to 3:27, and none of the hour hands is in a plausible location.

This "good-looking but implausible" attribute is characteristic of older art generation models, not just regarding clock dials.

Now we'll switch to OpenArt. GPT Image 2 and Seedream and some others are also available here, but OpenArt seems to specialize in rather off-the-wall specialty art models. The first shown here, which is also the newest, is Recraft V4, a specialty art model.

Recraft is also among the first to have good text rendering. It is also known for creativity; to a certain extent, one needs to be very specific with one's prompts, and it will still usually throw in unexpected elements.  For that reason, I use this one when I'm creating more whimsical images.

The times shown are all over the place, similarly to Lucid Origin, but are generally more plausible, as though Recraft has a better handle on the relationship between the minute hand and the hour hand. But check out the second dial, with 22 up there!

WanAI, part of Alibaba, produces the Wan family; this is Wan 2.7. They focus more on video generation, which requires better consistency from image to image. Come to think of it, it has been a good while since I worked with a model that allowed me to set the seed. Most models now keep the seed hidden.

This is also a model that will throw in extra stuff. The four central clocks all read 10:10, and all have the hour hand pointing right at the 10. The other clocks nearby have a variety of readings, and sometimes an extra hand or two. They are more "clock-ish", not really clocks. That makes this a model I might use to create a scene on another planet, where the idea of a "clock" only roughly approximates our own.

Grok Imagine Image 2.0 belongs to SpaceXAI. It aims to be good-to-excellent with everything.

Here, it's better than most of the others, but falls behind GPT Image 2. The four clock dials all have their minute hands aimed at :25, but the hour hands are stuck at the 3.

I do like the little motto at far right. I use this model when I'm trying out a prompt on several models, looking for creative variations.

The Chinese company Kuaishou Technology has the Kling art models; this one is Kling 3 Omni.

Here, the times shown appear random, and there is no relationship between the hour hands and the second hands; none is plausible.

Rather than the "beige wall", this appears to be a display board. The rest of the clock shop is a nice touch. This is another of the models I use to see what creative stuff it might throw in.

Here is the last experimental image. Alibaba also owns the Qwen art models; this is Qwen Image 3.0.

The times shown are approximations to 3:27. None has the hour hand in a proper location. Close, but no cigar.

Qwen models had an early focus on more accurate people, when DALL-E3 was still putting extra fingers and toes and other distortions on them.

This is the only image of the ten that shows prices for the clocks. It must be channeling a future time; the prices are about twice as high as similar clocks at most stores...at least the ones I shop at!

Clocks are one of the edge cases I use, firstly to judge the accuracy and comprehension of an art model, and secondly to detect a deepfake. In particular, I have seen a few AI-generated podcasts recently that showed the avatar in an office setting with a clock in the background. The clocks' hands didn't move.

For the interested: another "tell" is that AI generated images and videos in particular are sharper than "real" photography or videography. As shown above, creating bokeh isn't hard anymore. But the avatars are a little too well-defined.

Beasts in our yards, beasts in our minds

 kw: book reviews, nonfiction, animals, pests, animal behavior, sociology, attitudes, risks

It's funny. We have the words "vermin" and "pests" for animals we don't like, but no single-word terms for animals we like. We get by with "animal allies" and "domestic animals", plus "pets" for our most-loved. Most literature about animals either advises us how to get rid of the ones we don't want around, or idolizes certain "charismatic megafauna". There are clubs of bird watchers and butterfly collectors and wildlife photographers (even when the "wildlife" is found in your neighborhood). Could there be clubs for those who care about bats, or beetles, or (shudder) roaches?

The more citified we are, the more likely we are to fear or despise animals—other than the pets or "companion animals" we dote upon. Even some bird watchers call city pigeons "rats with feathers" and some of us go to great lengths to keep raccoons out of our garbage containers (Hint: If you have room for it, use a compost pile for everything that's even slightly decomposable).

Sometimes I've wondered, "What do animals think of us?" If you could converse with one, what would it have to say? Perhaps when a monkey sees a human, it thinks of her as a funny sort of hairless, tailless monkey. When a fish, even a shark, sees a human swimming, it probably thinks it is seeing a particularly clumsy seal. Perhaps the garter snake in the garden would advise you, "Don't spray the veggies. I don't like my insects covered in pesticides."

In her book Beauty of the Beasts: Rethinking Nature's Least Loved Animals, Jo Wimpenny asks us to look at animals differently, and in terms familiar to them. For example, when a moth follows you in through your front door, it isn't "invading". For most moths, everything inside your house is useless. It cares nothing for the furniture, the carpets, the lamps, the pictures on the wall (except as something to hide behind). If it is a female of one of the five or six species called "clothes moth", and she finds her way to a closet with some wool garments, if she thinks anything, it is, "Oh, goody! Oodles of stuff for my babies to eat," as she lays her eggs on your favorite sweater (Hint: woolens that you value are best kept in a cedar chest or cedar-lined closet, or even a chest freezer).

You might say, "You're just giving in to the beasts!" I am actually giving in to the math. There are about 11,000 species of moth so far known in the US, and more than 150,000 worldwide. Only about five have wool eating larvae. Most species outnumber humans thousands to one. There are probably several thousand moths the size of a thumbnail or smaller in my yard right now. In the evening, when they are airborne, getting into or out of the house must be done quickly, or at least one will get inside. If I can, I'll either catch it to release outside, or crush it. Probably half escape my attempts, but I can trust the house spiders to get most of them. And yes, we do have a cedar chest.

Ah, yes, the house spiders. Based on what I find in their webs, they mainly eat fruit flies and midges and the occasional housefly. Yesterday as I left the house, I felt a tiny thing land on my hand. I was thankful it was not a mosquito, but a newborn spider with a body no larger than a poppy seed. I blew it off into an Azalea bush, with a wish, "Catch lots of bugs for us." I knew it had several siblings still inside. I don't evict a spider until it is too big to stand on a nickel. Fortunately, my wife is not creeped out by having indoor spiders. That's good thing, because no house is free of them. Not one! They stay in odd corners, weaving tiny cobwebs, eating most of the tiny critters that get inside, and each other (population control!).

The book mentions that spiders are known to dream. I don't know how researchers figured that out! Can you imagine what they dream about?

The ones that were studied are jumping spiders (see this article in Harvard Gazette). While sleeping, they twitch their legs and their eyes move (I hadn't before known that spiders' eyes can move!). Just like REM in mammals and other animals.

Ms Wimpenny doesn't start with insects and spiders, though, she starts with fearsome critters, the ones that could eat us, or at least do a lot of harm: sharks and crocodiles. The movie Jaws set back both research and acceptance of sharks by decades. Earlier treatments (not mentioned in the book) by Jacques Cousteau that depicted sharks as mostly bloodthirsty set the stage for Jaws and the more recent Shark Week.

It's interesting that people fear sharks the most. I heard this joke when I was a graduate student:

Families are enjoying the beach at Cape Cod, when someone yells, "Shark!" Everyone runs out of the water, hops in the car, starts the engine to run the air conditioner, and has a smoke.

This was 50 years ago when almost everyone smoked, and the first "no smoking" restaurants were just getting started. Diseases caused by smoking kill half a million Americans yearly, and auto exhaust at that time probably killed half that many. The number of people killed by sharks in a typical year? Two.

The fact is, sharks are more social than we knew in the past. The author tells of social and even caring behaviors. Crucially, there are many species of shark, and they differ a lot. But even the "king of sharks", the great white shark, is a somewhat social species. They don't think of humans as prey. They probably don't think about us at all. Nudging and biting are how they gather information. The animal that does consider us prey is the crocodile. One Australian told the author that if someone jumps off a certain bridge and tries to swim just the 100 meters or so to shore, they won't make it. A saltwater crocodile will "take" them within a minute or less. But if you're in Australia and someone hollers, "Croc!", everyone cranes their neck for a look.

Bat-phobia is not as prevalent as shark-phobia, but is equally real. People think bats are creepy, even if they're too young to remember the spate of Dracula movies of the late Twentieth Century. We are generally slow-moving and bats probably think of us as furniture…furniture that is ill-equipped to catch food.

I spent half a summer doing geology in Nevada. There was a bright light on a tall pole, far from our bunkhouse. Moths and other insects swarmed around it. So did bats, catching them. Sometimes one of us would get a flat pebble and throw it up spinning into the light. A bat would dive-bomb it until it was about a foot or two away. Then its echolocation would be able to discriminate that the rock wasn't a moth, and the bat would swerve off. We were never able to induce a bat to capture a pebble.

Strangely, for a book intended to make us feel less fearful and more "friendly" towards animals, particularly the creepy or disgusting ones, the last chapter before an Epilogue tells of cute animals behaving badly. I won't go into it; much of it is gruesome or grotesque. Think abusive penguins and otters. Really abusive! The author's intent is to emphasize that animals don't live by our rules. I lived among rattlesnakes on that field trip mentioned above. I saw them every day. They were afraid of us humans. When I saw one it was usually retreating. Since it was summer, they never came inside; they like warmth. Had I been so foolish as to camp on the ground in cold weather, it's possible I might have awoken one morning with a snake curled up in my sleeping bag. I knew better!

Do you like the big cats? Many people do, but of course, at least in America, we enjoy them from the safety of a zoo enclosure. Where people and big cats live near one another, it's another story. Just in India, about 70 people are killed yearly by tigers. That's ten times as many as sharks kill, worldwide. This cartoon is based on one I saw sixty years ago (All these cartoons were produced using GPT Image 2 in the OpenArt site). 

There is an uphill battle to improve the public's perceptions of animals. AI will probably accentuate this, because LLM's like ChatGPT or Gemini have been trained on human text, much of which has outdated, inaccurate views of animals. Our literature is filled with over-cutified or overly villainous animals. Hardly any balanced depictions exist.

Other than humans, what is the most deadly animal? The mosquito. But out of thousands of species of mosquito, only a few carry diseases. Unfortunately, one of them is malaria, and others include West Nile encephalitis and chikungunya (bone-break fever). But not all mosquitoes need blood; many use plant sap. Of the bloodsuckers, few will utilize human blood. To get rid of malaria, however, getting rid of Anopheles mosquitoes isn't the right approach. Effective vaccines against the parasite the mosquitoes carry are just being released, after many years of research. I hope it will be cheap enough to allow near-eradication of the parasite. Meanwhile, mosquito nets are the fallback measure. Further, bats and many kinds of birds subsist on mosquitoes and other flying insects. Kill the insects and what will happen to them?

All of nature is sets of interlocking ecosystems. Cut one link, and it can shut down a whole chain, or hinder it for many years until it readjusts. We are seeing that all around us. I think the key to rewiring our attitude towards nature has two parts: (1) Recognize that we are part of nature, and (2) Discard the notion that any animals think like we do; they don't recognize boundaries that we hold inviolable.

Here's hoping that this book makes a difference!

The concept of image gamma

 kw: photo essays, lighting, gamma, image editing, low key, high key, key adjustment, gamma adjustment

I was asked to explain gamma in images. There is a mathematical explanation, but not many folks would have a gut-level understanding of the math. Instead I'll describe the concept: 

Gamma is the relationship between the brightness of the light at every point in a scene and the numbers used to store each point in a pixel and display it. In an image file, adjusting the gamma of the image raises or lowers these numbers to "brighten" or "darken" the image, while keeping the zero and 100% points fixed.

When the gamma equals 1.0 the relationship between light intensity and the stored values is a straight line. In such a case, when a pixel received half the amount of light compared to the brightest pixel in the image, the number stored for it is 0.5 times the maximum possible number. For most digital images, the numbers range from 0 to 255, so the midpoint is 127 (only integers are used, so the extra half a digit is dropped). When gamma is changed for an image to a number other than 1.0, the relationship becomes a curve. The result is an image that appears brighter or darker, as if the exposure had been different.

Adjusting gamma is different from changing the overall brightness. Raising "brightness" using most image editing software just adds the same amount to every pixel's value, and lowering it subtracts a fixed amount from every pixel's value. Just a little of this may be OK for some images, but a more nuanced approach works better.

This is the "Color Curves" tool in GIMP (a freeware program similar to Photoshop). Moving portions of the straight line that goes from lower left to upper right adjusts the relationship between the values in an initial image and those in an adjusted image.

Note that below the histogram the "Curve type" is Smooth. For simple changes, we can click and drag a point on the line up or down. The program obligingly adjusts the image so we can see the result in real time.

If we drag the center of the line upward, for example, the line changes to a curve passing through that point, showing the change caused by adjusting gamma. Dragging a point not at the center uses a different function, related to gamma but not the same.

The next image shows how the curve looks when the center point is raised a little.

The center point started with a value of 128, seen in the "Input...Output" indicators, and it now has been changed to 160.

The whole line was not shifted as a straight line, but is now a curve that passes smoothly from 0 to the midpoint to the upper right at 255.

In terms of gamma, this raises it from 1.0 to a higher amount. The effect on the picture is to brighten the midrange the most, with less and less brightening towards either end of the histogram.

Next I'll show how the red, green, and blue histograms are affected by this change, and then the image itself and its adjusted version. These histograms were produced using IrfanView:

The images I used are small, only 800 pixels across, so the "before" on the left is jittery, and the "after" on the right is positively seismic! In the portion of the histogram that was stretched the values were shifted, which "pulled them apart", leaving "holes" between. The overall shape of the histograms reveal that the values were all shifted to the right (brighter), stretching the leftward (darker) portion and squeezing the rightward portion. 

Next we see first the image I started with and then its adjusted version. The primary visual differences are that more detail is seen in the shadows, the whole shop appears a bit brighter, and there is a little more emphasis on the woman.

The original image was 1,376x768 pixels; these are 800x447. I used a rather short prompt: "The essence of ingenuity with a steampunk vibe", using the Google Flow site, which makes use of the Nano Banana family of art engines for still images. Unless the prompt includes a request for a bright image, or high key, this engine, like all the others I have used, tends to produce medium-to-high contrast, moderately low key (sorta dark) images. Sometimes the image can be quite dark, as we will see.

When I want to adjust gamma only to raise the key of an image, I edit with IrfanView, which has a control to set a numeric value between zero (totally black except those pixels that started out with a value of 255) to 10 (almost totally white). GIMP doesn't use numeric gamma so I don't know the exact value GIMP used for the adjustment I made, but it is probably about 1.2.

Now let's see how gamma adjustment affects images that were deliberately produced at either high key (lots of white) or low key (lots of dark gray and black).

Here is a common treatment of a baby picture: overexposed on a white background. The attached histogram shows that there are very few darker pixels. The slightly dimmer peak at the right is the shadow, barely visible to the left of the baby's head. The "baby pixels" make up the "mountain".

I used a gamma adjustment of 0.75 to bring out more color in the baby's face. Although the shadow is a bit more distinct, I like this picture better overall.

Next, here is a low key example. I have made a lot of images of imaginary caves. This is one of a series I call Spelunker's Paradise, based on a 15-word prompt. While cave imagery tends to be low key, this was the darkest of the bunch. The engine was Ideogram in Leonardo AI. The histogram has a big peak near "black" and not much else, and in the image one can see only hints of the scenery.

I use IrfanView with a gamma adjustment of 1.2 or 1.3 for moderately low key images. In this case more drastic treatment is needed; I first used 1.5:

A lot more is visible. Based on my own spelunking experience, this is probably realistic. However, I also tried a gamma adjustment of 2:

If I were going to print the image for hanging on my dining room wall, this is the version I would use. You can see in the histogram that all but a few of the darkest areas have now been pushed to be brighter than nearly all of the first image. If the gamma were to be raised even higher it would look unnatural.

In all the histograms of gamma-adjusted images, the jitter-to-zero shows that when RGB values were changed to be larger, gaps resulted. The human visual system is very forgiving of small glitches in color. If I wanted to produce an image with strongly adjusted gamma, and still have a smoother color histogram, I would upscale the image 2:1 using Upscayl, adjust gamma, then reduce the result again to the original size. The final image would have a smoother histogram and this would also potentially reduce image artifacts that could show up otherwise.

The key takeaway is that adjusting gamma is like changing the exposure after the fact.

He entertained a century

 kw: book reviews, nonfiction, memoirs, autobiographies, entertainers

Dick Van Dyke performed in his high school's a Capella choir and drama club until he left early to join the Air Force in 1944. This early experience began to lead him to a life of entertaining, and he is still at it, at age 100. By his own admission, he has slowed down quite a bit, but is still active with the Vantastix, a men's quartet he formed in 2000, and in occasional special events. It all adds up to more than eighty years, or my entire lifetime and then some!

I most fondly remember him as Bert in Mary Poppins. Being of a musical bent myself, I really liked his one-man band (This picture is the best sketch-style image I could produce with the NanoBanana 2 art engine).

At age 99, with every prospect of reaching 100 years and beyond, he wrote a third memoir: 100 Rules for Living to 100: An Optimist's Guide to a Happy Life.

He tells us at the beginning that the number of chapters is approximately 100, and not to hold him to exactness. I didn't count, but I did notice that the text is 309 pages, and the chapters seem to average 3 pages each, so it's pretty close. There is no Table of Contents, no Index, just the (approximately) 100 chapters of advice, anecdotes, and adventures.

He sets the tone with the first chapter, "Don't Act Your Age", which is great advice if you want to portray both Bert and the ancient banker in the same film, as he did. And it is generally healthy. He follows with "Make Your Own Rules", in which he explains why he offers more examples and less advice. That resonates with me, because I've found that if I offer advice, and it is followed, one possible outcome is that all goes well but the person I've advised will always feel less ownership of the result; and if it doesn't go well, I get the blame.

I'll just mention a couple of other items. "Tell Your Hardest Stories", beginning on p. 105 (and one of the longer chapters), shows how our experiences can help others who have similar problems, if both parties are willing to be vulnerable and truthfully relate them. He concludes, "Each of us has our own hard stories of crisis and struggle. When we hold them in, out of fear or shame, they control us. But when we tell our stories, we're in the driver's seat. …we are literally helping others to survive…"

And, "Write it Down" (p. 189) is exceptional advice for most of us. I've learned three ways we might relate to lists or written notes:

1. Some folks make lists and follow them.
2. Some don't make lists.
3. Some lose lists. I am in this category!

I've learned how to cope with being a list-loser since I got a smart phone: I have a Google Doc where I put lists I want to recall and random ideas before they get away. When an item is outdated I can remove it.

Dick Van Dyke has had many lucky breaks and much good fortune, but he hasn't been uniformly lucky, and he tells some of his hard stories. His claim to the title "optimist" is that he bounces back, sooner or later, and usually pretty soon. It is good to remember the past for the lessons learned and for the happy memories. It isn't as good to dwell on past sadnesses. He tells of a couple of those that took quite a while to resolve. I'll let you do your own reading to find them. They don't detract from a great fulness of happy stories.

It is a fun, engaging book. I was surprised at the end when he writes with great feeling about campaigning for Bernie Sanders in his three major Presidential bids. I guess everyone is allowed a bit of insanity… Both he and Sanders greatly benefited from the capitalist system that Sanders claims he wants to destroy.

Can science be rescued?

 kw: book reviews, nonfiction, science, history of science, scientists, biographies, inquisition, polemics

At the defense of my PdD dissertation, one prominent professor growled, "This isn't geology, it is computer science." To a certain extent, he was right. I was surprised that he was the one to say it, though, for among the faculty he was the most computer-literate. I had used a great deal of computer modeling and simulation in conjunction with my research. This was 1983, and the geology faculty in general were much more comfortable out in the field with a rock hammer than they were poking at a computer terminal connected to a mainframe computer hidden away in the basement of the Electrical Engineering building. Desktop personal computers were rare. On the other hand, I was quite computer literate, and I was at that time an adjunct professor of computer science.

That wasn't the only objection to my work. I'll discuss that in a moment. The turning point from interview to inquisition began when someone asked, "Do you really believe this stuff?" I answered, "Believe? That's the wrong word. 'This stuff' cannot save my soul. I believe in Jesus. But 'this stuff' is the best scientific explanation for the phenomena we observe." Hostility ensued. I felt like St. Sebastian, who was tied to a tree where archers used him for target practice.

I had spent a few years of research to determine what effect directional heat flow in Earth's crust might have on the emplacement of "gneiss domes" such as the Black Hills of South Dakota. My supervising professor and I went all over the Black Hills to gather specimens of rocks that had originally underlain the domed structure and rocks in the lowest layers above the granitic/gneiss core of the Hills.

I needed to determine the thermal conductivity of these rocks in three directions. I used a large core drill to cut out cylinders, which I sliced into sections a centimeter or so in thickness, and then I polished the flat surfaces for use in a special press with heaters and thermocouples. I still have some of these. I call them my "hard disks".

The ratio of heat flow across the grain, versus along the grain, is the thermal anisotropy. For a pure, platy mineral such as mica, I found that this ratio could be as high as 6:1. For the shales and schists in the lower sections of the Black Hills, I measured thermal anisotropy between 1.2:1 and 1.6:1. The initial discussions of the early results with my committee were not encouraging. I did a series of computer simulations which showed that the excess heating caused just by thermal anisotropy was no more than a few degrees.

As it happened, I spent time that summer on a project about fluid flow. Another graduate student and I found that the anisotropy of fluid flow in layered rocks could be quite high, 10:1 to 100:1 and even greater. Any fluid flowing through these rocks would be very strongly affected by this directional fabric, and the solid-rock thermal anisotropy wouldn't matter at all. As a side project, the student and I dug into literature and studies about fluid flow in the crust. We determined that even in "dry rock" there is water, and it is in motion, and if there is any heat flow it is directed almost entirely by fluid flow. This was new; there was no literature on this particular point.

Therefore, I wrote a dissertation in which I stated that my initial theory was not tenable, and that further study would be needed to quantify the results of anisotropy in fluid flow. This was quite unpopular and it apparently stepped on a few toes. It was "this stuff" that caused my committee along with the other gathered faculty to reject my dissertation. I became, not a PhD, but AbD, "all but dissertation". Coda: I had an established relationship with scientists at a national laboratory in California. I passed my materials on to them, and they were glad to have them.

With such a background, I was well primed to receive Matt Kaplan's historical survey of scientific mavericks in I Told you So! Scientists Who Were Ridiculed and Imprisoned for Being Right. 

The book is in part a biography of Ignaz Semmelweis, the best-known of those who learned how necessary it is to clean a doctor's hands and equipment before treating a patient. Semmelweis was also the one who seems to have suffered the most for it. The various parts of his life story are woven with others. The book opens with a scene in which a graduate student named Alison, at a poster session, is being shouted into submission by a gaggle of offended professors. Her sin? Being right. Theirs? Feeling threatened by the truth. It costs something to learn something new that reveals the flaws in older ideas.

One scientist who didn't fall prey to this phenomenon was Louis Pasteur. He had the political acumen to make sure he had powerful friends when needed, he was a showman who used public demonstrations (thoroughly prepared beforehand) to publicize his ideas, and he was careful to tailor his presentations to make it seem like his discoveries came as flashes of insight, where in reality he labored long and hard upon them.

The author tells stories of others, including friends and acquaintances, who survived being right only because of more established mentors, and by finally "being right" in a sufficiently public way that could not be denied. Katalin (Kati) Karikó and Drew Weissman received the Nobel Prize in 2023 for working out the therapeutic use of mRNA agents, which led to the "vaccines" used during the Covid-19 pandemic. But Kati had rough going for many years, suffering abusive supervisors and disdainful department heads. Somehow, she plowed through, and collaborations such as that with Dr. Weissman kept her work going. Many others have been less fortunate. Alison left science, as have several others mentioned in the book.

Being right is not enough to ensure being received. Max Planck is known for saying that science proceeds one funeral at a time. My father used to speak of the "Moses Effect", meaning that it takes 40 years in the wilderness for a generation to die out so a new generation can make progress.

Semmelweis seems to have had no such advantages, and many disadvantages. He was not the only doctor to learn the value of chlorine-laced wash water for disinfection. But in his own hospital, he seems to have had the knack for presenting his findings in a way that didn't just challenge other doctors, including his supervisor, he actually indicted them along with himself for killing thousands of young mothers through ignorance.

Arthur C. Clarke's "first law of science" is, "When a distinguished but elderly scientist states that something is possible, he is almost certainly right. When he states that something is impossible, he is very probably wrong." The trouble is, the distinguished and elderly scientists usually stopped having good ideas half a lifetime ago and feel threatened by the next generation.

These days, it may seem we have moved beyond the times of Galileo, who had to stand before the Holy Office of the Inquisition for saying that Earth moves. Scientists these days don't have the rack or a team of archers to deal with mavericks. But they do have powerful weapons to destroy us anyway. Today, money runs science, in a way much more pervasive than ever before. The "old guard" controls the flow of money, which determines who can afford to do experiments, and who must find a different profession.

I think of two doctors I knew well. One, named Henry, was a radiation oncologist. He once said he could cure most cases of breast cancer, at lower cost than chemotherapy. However, most patients get steered toward chemo rather than radiation. When I asked why, he just rubbed two fingers together, the sign for "money". The other doctor, named Benjamin, is a friend I once asked about how thyroid issues affect fertility. He said, "One third of infertility cases are due to low thyroid, but it is the last thing tested, when it is tested at all." Of course I asked why, and he said, "Treating low thyroid is the cheapest remedy."

In the last chapter the author discusses possible remedies for the prejudice-dominated grant-awarding process. While he has some good ideas, including ideas that are being used on a small scale, he recognizes that government action would be needed to motivate widespread change. Given the huge pharmaceutical and medical lobbying that goes on, useful change is unlikely. I fear that Max Planck was only partly correct, because he said "one funeral at a time." It takes several.

I think of Ignaz Semmelweis and the other pilloried scientists as being like Jeremiah: never wrong; never believed.