jump to navigation

Artificial Intelligence Finds an Antibiotic February 20, 2020

Posted by stuffilikenet in Applications, Awesome, Geek Stuff, Science, Star Trek Technology.
add a comment

In a news flash from M.I.T. scientists announce the discover of a (potentially) low-toxicity molecule which interferes with bacteriological cell walls’ ability to maintain electrochemical gradients, which are essential to creating ATP, the main energy molecule in, well, everything: the cells would starve. An A.I. was trained on 2,500 molecules and then scanned the Broad Institute’s Drug Repurposing Hub, a library of about 6,000 compounds. The model picked out one molecule that was predicted to have strong antibacterial activity and had a chemical structure different from any existing antibiotics. Using a different machine-learning model, the researchers also showed that this molecule would likely have low toxicity to human cells.

It worked very well in vitro and in mouse models on a bunch of stubborn microbes that are getting to be pretty resistant to everything we have: Clostridium difficileAcinetobacter baumannii, and Mycobacterium tuberculosis. The drug apparently worked on EVERYTHING they tested, except Pseudomonas aeruginosa.

We Are Legion (We Are Bob) February 18, 2020

Posted by stuffilikenet in Awesome, Books, Brilliant words, Geek Stuff.
add a comment

What happens when you take a slightly-brighter-than-average engineer (Bob) and give him immortality in a starship equipped with extremely advance prototyping machines, then  tell him to set up infrastructure for colonists to follow in a few decades?

In the fertile imagination of Dennis Taylor, Bob takes over guardianship of not only the whole human race but at least two other sentient species, invents FTL communication, planet-movers and nifty full-sense android bodies. It takes  quite a few decades, but Bob has made many. many clones of himself, and they all have all the time in, well the universe with which to foil the Evil Plans of man, machine and alien.

Well read by Ray Porter with excellent inflection and pacing, all three Bob books skip right along, neatly compressing the decades into digestible chunks and holding a listener’s attention well enough to keep me awake during six-hour drives that end at 2AM…like last night.[0]

The dialogue is interesting, the characters internally consistent and the technology descriptions are pretty darn good. Taylor is very obviously a sci-fi fan and geek, and we should all be glad for this: it lends credibility to his character’s engineering comments and descriptions of space, the choices of star systems (he apparently did some homework) to visit, and the tropes he chooses to infuse with credibility.

In the immortal words of Joe Bob Briggs, check it out.

 

[0] Goddamnit.

 

Where I Work December 27, 2019

Posted by stuffilikenet in Awesome, Geek Stuff, Toys.
add a comment

we are a hard-workin’ bunch, let me tell you:

A.I. and Gene Regulation December 27, 2019

Posted by stuffilikenet in Applications, Brain, Geek Stuff, Mutants, Science.
add a comment

Understanding gene regulation is a bitch. Seriously, this is one of the thornier problems of science today and it is because the complexities of living cells, with the thousands of proteins in each cell, make tracing a single protein’s regulation just as complex as hell. Smart guys Tareen and Kinney have figured out a way for AI to interpret (some) interactions using massively-parallel reporter assays to figure out the biophysical basis for (some) gene regulation…which is more than we have had heretofore. They did this by assigning nodes and weights with explicit physiochemical interpretations. This last is the important bit; many AI algorithms are very difficult to interpret, so the underlying “logic” is impenetrable to humans.  The smart guys made many of the decisions explicit, so they would be better able to understand the “logic” by which the characterizations were derived.

 

 

Homework: Biophysical models of cis-regulation as interpretable neural networks, Ammar Tareen, Justin B. Kinney BioRxiv,

Multi-dimensional Blood Testing and A.I. December 23, 2019

Posted by stuffilikenet in Applications, Awesome, Brain, Brilliant words, Geek Stuff, Science, Star Trek Technology.
add a comment

I suggested long ago that sufficiently-comprehensive blood tests could effectively predict a person’s risk of developing a broad array of different diseases. We would use artificial intelligence to find patterns of varying concentrations of blood proteins to predict and/or diagnose disease. Someone much better funded than me has a newly developed platform called SomaScan which can scan five thousand individual proteins from a single blood sample.

In a new study testing the efficacy of predicting 11 different health indicators using these protein expression patterns some models were much more effective than others, such as the protein expression model predicting percentage body fat. The cardiovascular risk model was cited as only modestly predictive, however, the researchers do suggest the protein-pattern-based system is generally more convenient, and cheaper, than many traditional tests currently available for evaluating health conditions.

The study in Nature Medicine was funded by SomaLogic which owns SomaScan, so grain of salt, people. But it’s exciting to see that someone is actually looking into what I feel will be the method of the future for maximizing health…also, the study used ~85 million protein measurements in 16,894 participants, which is a pretty damn good sample size.  Plenty of data there for an A.I. to examine for hidden relationships.

Homework:

Plasma protein patterns as comprehensive indicators of health, Nature Medicine, Stephen A. Williams, Mika Kivimaki, Claudia Langenberg, Aroon D. Hingorani, J. P. Casas, Claude Bouchard, Christian Jonasson, Mark A. Sarzynski, Martin J. Shipley, Leigh Alexander, Jessica Ash, Tim Bauer, Jessica Chadwick, Gargi Datta, Robert Kirk DeLisle, Yolanda Hagar, Michael Hinterberg, Rachel Ostroff, Sophie Weiss, Peter Ganz & Nicholas J. Wareham

Enchanting Furby Toy Hack December 20, 2019

Posted by stuffilikenet in Applications, Awesome, Geek Stuff, Toys.
add a comment

I do wish I had thought of this.

Moore’s Law, and Progress December 11, 2019

Posted by stuffilikenet in Awesome, Brilliant words, Geek Stuff, Video.
add a comment

An Accelerated Pace of Change

Moore’s Law has translated into a faster rate of change for society as a whole.

A new idea, like the smartphone, can get immediate traction because of instantaneous communication, increased global connectivity, and the ubiquity of information. New tech advancements can now change business or culture in a heartbeat:

The accelerating rate of technology adoption

Further, since software is a “layer” built upon the foundation of computing, it means that digital products can be replicated at almost no marginal cost. This is why a phenomenon like Pokémon Go was able to captivate 50 million users in just 19 days.

Imagine this kind of scalability, when applied to things like artificial intelligence or virtual reality.

–stolen freely from Visual Capitalist. I’m sure they’ll ask me to take it down soon. But look at the possibilities, people!

Capable Modular Robots November 9, 2019

Posted by stuffilikenet in Applications, Awesome, Geek Stuff, Science, Star Trek Technology, Toys, Video.
add a comment

 MIT’s Computer Science and Artificial Laboratory’s M-Block robots can self-assemble into different structures. The little cubes (in development over the last six years) have gained the ability to jump, flip, spin, and recognize each other. A barcode-like system on each face of each cube, allowing them to identify the other cubes around them.

The engineers wanted to see if the M-Blocks could (for example) form a straight line or form a random structure using the new communication algorithms. They waited to see if the blocks could determine how they were connected, and then what direction they would need to move to create that line. They found that 90% of the block swarm knew which motion and guidance to move to accomplish the task. I’m curious to know what the other 10% did…

Engineers hope to create a more substantial swarm of blocks (>16) that can assemble to form more complex structures with new capabilities.

3D Scanning to Print (Photogrammetry) October 8, 2019

Posted by stuffilikenet in 3D Printing, Applications, Awesome, Geek Stuff, Photography, Publishing Tools, Star Trek Technology.
add a comment

Due to an unfortunate incident while moving from the ver parted lips of Hell (Greenville, CA) I have had to replace a few parts on my Casio CDP-120 piano keyboard. Basically, the highest key (C for those who need to know) was/is smashed and needs a replacement. Aha! I have this iPrusa printer and can whip up a new key post-haste using
Fusion360 to measure and design a new key.  Right?

No. It’s a part with surprisingly complex geometry, so I thought I would try photogrammetry to create a model and see how that works.  I never tried this before, so what could it hurt? Can’t be any harder than Fusion360 to master. Right?

No. There seems to be no (free as in beer) software which can do this which is simple enough for me to install (grumble, grumble) on this company Mac, and I’m not sure company policies built into the security suite on this company Mac would permit it (grumble, grumble) anyway. Since my replacement Windows box isn’t here yet (grumble, grumble) I have to pay for a cloud-based solution…Altizure.com.  So I took 56 pictures of the C key an octave down with my cellphone camera, submitted them to Altizure.com‘s loving embrace and was rewarded with a fairly competent render of my key:

Screen Shot 2019-10-08 at 11.06.36 AM.png

I say “fairly competent” because it has lots of hole in the sides.  No matter; I can probably fix it in software modify it in Fusion360 to fill in the holes.  I will likely need to use Fusion360 to get the dimensions useably correct in any case.

Now I have to steel myself to pay the subscription rate of something like $70 per project. It would be worth it if I was a company making a prototype with great prospects for the future…but I’m not (grumble, grumble).

The key costs about five bucks from Casio’s parts subcontractor, plus six or so to ship it. I know which I am going to choose.

It’s a shame, really; I was looking forward to monkeying around with it. Maybe next time.

However, I don’t give up easily; I got 3DF Zephyr and used the (limited to) 50 pictures in it at the highest defaults in the wizard to create this render:

c-key-3df zephyr

My army of loyal and discerning readers may notice a bit of degradation here and there…yeah.  I will try again, but this took an hour or so and was NOT automatically rendered, but went through stages. In all fairness it would have been shorter if my graphics card had CUDA…but I don’t have a graphics card, just whatever Lenovo thought was cheap but still enough to get by on. Still, it needs work, and it’s midnight.  Possibly updates to follow, if I get any sleep.

EXCITING UPDATE:

I got a better render with 50 pictures and the most extreme settings I could figure out:

c-3df-zephyr-2-extra tasty

It’s got the holes that Altizure’s render had, but 3DF Zephyr is free to use.  Now I have to figure out how to make the save-able form of this (an .obj file) into an .slt and thence to gcode. With luck, I will have a hideous key to use before my real one arrives.

 

 

 

Giggling Maniac with Gatling Gun July 16, 2019

Posted by stuffilikenet in 3D Printing, Awesome, Geek Stuff, Toys, Uncategorizable, Uncategorized.
add a comment

It’s lovely what you can do with 3D printers. It is, of course, best illustrated by a cackling madman with a Gatling Gun.

 

EXCITING UPDATE: I tend to forget the excellent work done in a related space by Jeorg Sprave, a happy German fellow who specializes in rubber-driven devices for impelling various projectiles:

As the video shows, he is also a giggling maniac, just this time with an electric fully-automatic crossbow.

3D Printing Progress May 10, 2019

Posted by stuffilikenet in 3D Printing, Applications, Awesome, Geek Stuff.
add a comment

There’s going to be a lot of that, now that critical mass of commercial systems are available. new materials are being used in additive manufacturing and new devices have considerably expanded the capabilities of systems in terms of speed, build volume and finish. It’s an interesting moment in engineering history, and nobody knows where it will lead.

Printing with [anything besides plastics] is fraught with difficulty, so interesting methods have been tried for substances like metals, clay, frosting(!) with varying success. Two methods have lately shown promise in metal and glass(amazingly enough).

First, metals. The most common method of depositing metals has been to embed the metal in something a bit more fluid, like in an ink suspension. This has the usual effect of having poor mechanical adhesion, because after the fluid dries the metal may adhere to itself poorly (likely) and there may be fluid contamination trapped in the metal layers (very likely).  Researchers got around this with an entirely new method, using a sacrificial electrode to generate ions of the metal and spraying those ions electrostatically. You can get insanely small resolution using this technique:

…and you can print with more than one metal by building both into the tip and just switching voltage from one electrode to the other:

Elegant as hell, isn’t it?

Then, glass: a team in France using chalcogenide glass (which softens at a relatively low temperature compared to other glass) produced chalcogenide glass filaments with dimensions similar to the commercial plastic filaments normally used with the 3-D printer. The research team then increased the maximum extruding temperature of a commercial 3-D printer from around 260 °C to 330 °C.  The result is pretty interesting:

An interesting proof-of-concept piece, this points to novel uses for chalcogenide glass commonly used to make optical components that operate at mid-infrared wavelengths. It’s not likely to be used elsewhere, as it’s a “soft” glass, but the feat is going to be useful in optics fabrication. Also, there are some low-temperature metal alloys that could probably benefit from this technique.

 

Homework: Alain Reiser et al. Multi-metal electrohydrodynamic redox 3D printing at the submicron scale, Nature Communications (2019). DOI: 10.1038/s41467-019-09827-1

E. Baudet et al, 3D-printing of arsenic sulfide chalcogenide glasses, Optical Materials Express (2019). DOI: 10.1364/OME.9.002307

40Hz Light Pulses Stop Alzheimer’s May 10, 2019

Posted by stuffilikenet in Awesome, Brain, Geek Stuff, Science, Star Trek Technology.
add a comment

Neuroscientists at MIT have published a paper which demonstrates that 40Hz pulses of light can somehow inhibit the progress of neurodegeneration in mouse model. This study is designed to figure out how a flickering light could stifle cognitive decline, using two unique mouse models engineered to overproduced the toxic proteins that contribute to neurodegeneration. The animals were exposed to light flickering at 40 Hz for one hour every day for between three and six weeks. It worked a treat;  mice engineered to overproduce tau proteins (that usually cause neurodegeneration) displayed no neuronal degeneration after three weeks of treatment compared to a control group that displayed nearly 20 percent total neuronal loss. The other mouse model, engineered to produce a neurodegenerative protein called p25, displayed no neurodegeneration whatsoever during the entire six weeks of treatment.

 

mousebrains.jpeg

The researchers then zoomed in on the light-treated animal’s neurons and microglia to study whether the treatment induced any unusual changes in gene expression. The light-treated mice revealed increased neuronal expression of genes associated with synaptic function and DNA repair. In microglia, the brain’s immune cells, there was a decrease in genes associated with inflammation.

Nobody understands how a 40 Hz flickering light can trigger these specific changes to gene expression deep in the brain, but human trials testing the sound and light treatment in Alzheimer’s patients have already begun.

Note: Adding a 40 Hz auditory tone to the process improved the efficacy of this treatment.  Your elderly parents can benefit from this by using gnuaural, an open-source generator of binaural beats for meditation and other psychological effects.

Homework: https://www.cell.com/neuron/fulltext/S0896-6273(19)30346-0

Crisis Management Down Under November 8, 2018

Posted by stuffilikenet in Brilliant words, Geek Stuff, Science.
add a comment

"Where have all the bloody teaspoons gone?" is an age old question in the workplace. In an article in the BMJ [not concerned with scatology, but British Medicine] from 2005, researchers at the Burnet Institute in Australia attempt to measure the phenomenon of teaspoon loss and its effect on office life. They purchased and discreetly numbered 70 stainless steel teaspoons (54 of standard quality and 16 of higher quality). The teaspoons were placed in tearooms around the institute and were counted weekly over five months. After five months, staff were told about the research project and asked to complete a brief anonymous questionnaire about their attitudes towards and knowledge of teaspoons and teaspoon theft.

During the study, 56 (80%) of the 70 teaspoons disappeared. The half life of the teaspoons was 81 days (that is, half had disappeared permanently after that time). The half life of teaspoons in communal tearooms (42 days) was significantly shorter than those in rooms linked to particular research groups (77 days). The rate of loss was not influenced by the teaspoons’ value and the overall incidence of teaspoon loss was 360.62 per 100 teaspoon years. At this rate, an estimated 250 teaspoons would need to be purchased annually to maintain a workable population of 70 teaspoons, say the authors.

The questionnaire showed that most employees (73%) were dissatisfied with teaspoon coverage in the institute, suggesting that teaspoons are an essential part of office life. The rapid rate of teaspoon loss shows that their availability (and therefore office life) is under constant assault.

One possible explanation for the phenomenon is resistentialism (the theory that inanimate objects have a natural aversion to humans), they write. This is supported by the fact that people have little or no control over teaspoon migration.

Given the widely applicable nature of these results, they suggest that the development of effective control measures against the loss of teaspoons should be a research priority

Hilarious. But wait; there’s more.

Exasperated by the disappearance, the scientists decided they would measure the phenomenon. Do the teaspoons really disappear over time? The answer was a resounding yes: spoons in research institute tearooms seem to have legs. While good fun, the research is a good example of a study design referred to as "longitudinal".

A longitudinal study uses continuous or repeated measures to follow particular individuals – in this case, teaspoons – over prolonged periods of time. The studies are generally observational in nature: the scientists simply watch and collect data over time. Typically, no external influence is applied during the course of the study. Beyond just working out where all the teaspoons have gone, this study type is also useful for evaluating the relationship between risk factors and the development of disease (for example, heart disease), and the outcomes of treatments over different lengths of time. In this study, the main questions posed by our researchers were to determine the overall rate of loss of teaspoons, and to work out how long it took for teaspoons to go missing.

They purchased 70 teaspoons (16 of which were of higher quality), each one discretely numbered and then distributed throughout the institute. Counts of the teaspoons were carried out weekly for two months, then fortnightly for a further three months. Desktops and other immediately visible surfaces were also scanned for "misplaced" spoons. After five months of covert research, the study was revealed to the institute, and staff were asked to return or anonymously report any marked teaspoons which may have found their way into desk draws or homes.

Good study design

This type of data collection provides a simple example of what makes a good longitudinal study. If we break it down, a longitudinal study needs to:

  • take place over a prolonged period (this study was done over 5 months)
  • be observational in nature (teaspoons were observed and counted, there was no intervention)
  • conducted without external influences (teaspoon users/thieves were not aware they were being studied until the conclusion of the study itself).

Results

 

The results show that 56 (80%) of the 70 teaspoons disappeared during the study, and that the half life of the teaspoons was 81 days (that is, half had disappeared permanently after that time). The study also showed the half life of teaspoons in communal tearooms (42 days) was significantly shorter than for those in research group specific tearooms (77 days). The rate of loss was not influenced by the teaspoons’ value. All of these pieces of information directly answer the main question posed by the researchers.

Conclusions

A longitudinal study is terrific at following individuals or teaspoons over a period of time and observing outcomes. But, by definition, the design means there can be no intervention (as we are just observing a phenomenon). The researchers could not employ a tool or an intervention to prevent spoons from being "misplaced", and the researchers could only report a spoon missing. As the study is observational only, there is no way of finding out what has happened to the spoon, just that it is lost. The authors were able to conclude that the loss of workplace teaspoons was rapid, and their availability in the tearoom was constantly under threat.

Homework: Megan S C Lim et al. The case of the disappearing teaspoons: longitudinal cohort study of the displacement of teaspoons in an Australian research institute, BMJ (2005). DOI: 10.1136/bmj.331.7531.1498

Package Installers for Windows and MacOS November 6, 2018

Posted by stuffilikenet in Applications, Geek Stuff, Publishing Tools, Video.
add a comment

I have been forced at gunpoint to use a Mac for the last six weeks at my newest place of employment, and not without a few tears. I had to learn to install IntelliJ, NetBeans and Eclipse (already had that one) for MacOS. The company which enslaves me uses MacOS’ Self Service app, from which I installed Homebrew. Homebrew does every installation you could possibly desire (well, nearly) and I installed in short order git, gradle, Java and IntelliJ–all correctly and findably by each other, managing the pathname (or whatever they are called in MacOS). I must say, this makes first-day setup for the engineers much quicker, and much simpler. Good thing too, since the poor sods are going to be working with a bewildering variety of the manifold technologies which enable the hydra-headed beast which is my employer.

It turns out that Homebrew is a MacOS-only product; but there are several package installers which can work with Windows, such as Scoop,

 

Chocolatey and Npackd,  I quite liked Scoop (hence the Youtubery), but you may wish to try the others.  Good luck; for your more complex setups this can be a real timesaver.

Fit Any Scatter Plot June 7, 2018

Posted by stuffilikenet in Brilliant words, Geek Stuff, Science.
add a comment

A wonderful paper in the archives of the University of Rochester  shows how any random scatter plot can be fit to a curve with enough parameters, and thence a lower number of same is often thought to be a good measure of an expression’s fitness for use…until now. “The mathematician John von Neumann famously admonished that with four free parameters he could make an elephant, and with five he could make it wiggle its trunk…The aim of this short note is to show that, in fact, very simple, elementary models exist that are capable of fitting arbitrarily many points to an arbitrary precision using only a single real-valued parameter θ. This is not always due to severe pathologies—one such model, studied here, is infinitely continuously differentiable as a function of θ. The existence of this model has implications for statistical model comparison, and shows that great care must be taken in machine learning efforts to discover equations from data since some simple models can fit any data set arbitrarily well.”

Tall claim?  Nope.  The author, Steven T. Piantadosi, shows two examples of data points fitted with a simple equation

 

image

can be fit to any arbitrary set of data plots……like these:

image

Mind you, the parameter θ needs to be calculated precisely: ”Both use r = 8 and require hundreds to thousands of digits of precision in θ.”.

Gee whiz (and hilarity) aside, the paper demonstrates the fallacy of using unreasonable models for this sort of algorithmic from-data derivation to create meaning from what might be noise, or Joan Miro’s signature.

Absolutely Juvenile July 26, 2017

Posted by stuffilikenet in Applications, Awesome, Geek Stuff, Mutants, Toys, Uncategorizable, Video.
add a comment

I enjoyed this immensely, but it is about one in the morning and I have been up since about five cleaning house and preparing for my next Great Adventure.

 

I am probably punchy.

Training Neural Networks to Write Bach in a Day! March 24, 2017

Posted by stuffilikenet in Applications, Awesome, Brain, Geek Stuff, Toys, Video.
add a comment

Fascinating bit of video here as Our Hero (not me, in this case) takes Bach (and later Mozart) MIDI files, creates an 88-character ASCII-character alphabet from them and trains a Recurrent Neural Network to output similar sequences.

The results (and a lot of the process) is shown in the video above.  Take your time and watch the whole thing; I wonder how long he would have to train the RNN to start outputting Baroque Muzak continually?

Why Trump? February 20, 2017

Posted by stuffilikenet in Awesome, Brilliant words, Geek Stuff.
add a comment

https://medium.com/@DaleBeran/4chan-the-skeleton-key-to-the-rise-of-trump-624e7cb798cb#.jmnimypj1 contains the answer to that puzzle.  Be warned: it is a long, thoughtful piece with some uncomfortable ideas for both left and right alike (but not alt-right).

HR 8799 Image Animated With Hot Jupiters January 28, 2017

Posted by stuffilikenet in Awesome, Geek Stuff, Photography, Science.
add a comment

Scientists have detected thousands of exoplanets in recent years, by watching varying brightness as they occlude their stars. This animation comes from direct imaging methods (not radio telescopes), meaning that the telescopes saw the Jupiter-sized planets directly (hot Jupiters are young planets that still glow in the infrared portion of the spectrum).

Jason Wang at the University of California, Berkeley has combined several observations of HR 8799 into the delightful GIF below. This is years of optical data, folks. “In this video you’re seeing real data,” he told Gizmodo1 of the video above. “I smoothed out the orbits so that it’s as if we’re watching [the planets] constantly in real time.”

orbitin beta pictoris hot jupiters

 

1 I stole the picture from Giz.  I feel no shame at all.  Follow that link, though; there’s much more!

Getting There, One Piece at a Time January 24, 2017

Posted by stuffilikenet in Awesome, Geek Stuff, Star Trek Technology, Toys.
add a comment

Open the Alexa app on your phone.

Use the More icon (the hamburger) in the upper left corner.

Select Settings.

Select the device whose wake word you wish to change.

Tap the Wake Word setting.

Select Computer from the drop-down.

You are done.

I would like to point out that Star Trek has so influenced culture that the United Federation of Planets is likely to happen any time now…we just need more planets.  We already have Majel Barret’s voice phonemes.  Now we need Google to sync up, a bit better response to voice meaning (the voice vector thing should help), a truthiness evaluator and bingo! Star Trek in your phone/home/office/laboratory/dungeon/whatever.