Archive for the 'Technology' Category

Jul 25 2016

A Tougher Turing Test

exmachinsert5In 1950 Alan Turing, as a thought experiment, considered a test for telling the difference between a human and an artificial intelligence (AI). If a person had an extensive conversation with the AI and could not tell them apart from a real person, then that would be a good indication that the AI had human-like intelligence.

This process became known as the Turing Test, and every year various groups administer their version of the Turing Test to AI contestants. The test has limits, however, and is generally considered to be too easy. It is also dependent on the skills of the human questioner.

Parsing Language

A recent AI contest used a different approach, the Winograd Schema Challenge (WSC).This is one of many alternatives to the Turing Test that are being explored. Here is the format of the challenge:

  1. Two entities or sets of entities, not necessarily people or sentient beings, are mentioned in the sentences by noun phrases.
  2. A pronoun or possessive adjective is used to reference one of the parties (of the right sort so it can refer to either party).
  3. The question involves determining the referent of the pronoun.
  4. There is a special word that is mentioned in the sentence and possibly the question. When replaced with an alternate word, the answer changes although the question still makes sense (e.g., in the above examples, “big” can be changed to “small”; “feared” can be changed to “advocated”.)

Continue Reading »

14 responses so far

Jul 15 2016

Virtual Reality, Augmented Reality, and Pokemon Go

Published by under Technology

PokemonGoThis past week my daughters were running around the neighborhood with their iPhones trying to capture virtual creatures. They weren’t alone – Pokemon Go became an instant sensation, becoming the most downloaded game app in US history. It will probably not last long (the app is somewhat buggy and the execution has issues), but I think it gives us a glimpse of the near future.

Predicting technology is tricky because it is difficult to know how people will receive and adapt to the technology. Until the tech is in the hands of the public, you cannot predict how popular it will be, and even how it will be used. Who would have predicted that texting would be more popular than video chat?

Well, the real story here is that Pokemon Go is perhaps the first time that augmented reality was put in the hands of the public, and they loved it (at least in the short term). This may prove to be an infatuation, but we’ll just have to wait and see.

Virtual Reality vs Augmented Reality

There is no question that these technologies are here and on the verge of becoming widespread. We are at that point similar to just before the iPhone, when smartphones were ready to become ubiquitous.

Continue Reading »

17 responses so far

Jun 27 2016

Senate Passes GMO Labeling Compromise

“You’re unhappy. I’m unhappy too. Have you heard of Henry Clay? He was the Great Compromiser. A good compromise is when both parties are dissatisfied, and I think that’s what we have here.”

– Larry DavidGMO_labeling-thumb

Senate Democrats and Republicans have reached a compromise on the issue of mandatory GMO labels. I am not happy with the outcome, but it could have been worse. Apparently pro-labeling advocates are unhappy too.

Last year the House passed a bill that would preempt mandatory labels. That bill stated:

(Sec. 101) This bill amends the Federal Food, Drug, and Cosmetic Act to require the Food and Drug Administration (FDA) to continue the voluntary consultation process established under the FDA’s “Statement of Policy: Foods Derived from New Plant Varieties.” In that process, the FDA evaluates a scientific and regulatory assessment provided by the developer of a food produced from, containing, or consisting of a plant that is a genetically engineered organism (GMO).

The FDA may require a GMO food to have a label that informs consumers of a material difference between the GMO food and a comparable food if the disclosure is necessary to protect public health and safety or to prevent the label from being false or misleading. The use of a GMO does not, by itself, constitute a material difference.

This was a reasonable bill, which acknowledges that the way a food is produced saying nothing directly about the safety or other properties of the end product.  A GMO can be perfectly safe, and a non-GMO can be unsafe.

Continue Reading »

86 responses so far

Jun 24 2016

Autonomous Cars Sacrificing Their Passengers

Published by under Technology

AVsI wrote last year about one aspect of autonomous cars (cars that are capable of driving themselves – autoautos, perhaps) that you might not think of right away but becomes obvious once pointed out: the cars will have to have an algorithm to determine what they do in an emergency situation.

For example, let’s say you are traveling down a road at 40mph and a group of people step out onto the road. The car’s computer calculates that it does not have enough space to come to a stop, so its only option to avoiding hitting the pedestrians is to swerve to the side, but there are obstacles on both sides and therefore swerving will result in a crash which might kill you. What should the car be programmed to do in this situation?

A recent series of surveys explores public opinion about autonomous vehicles (AVs).

To simplify the matter a bit, we can refer to such cars as utilitarian, meaning that their algorithms are designed to minimize the total loss of life without giving priority to the car’s own passengers. You can imagine a number of specific scenarios, but that is the essence of the issue. Should the car minimize risk to its passengers, or to all people equally regardless of their relative position to the car (pedestrian or passenger in another vehicle)?

Continue Reading »

27 responses so far

Jun 03 2016

Human Genome Project Write

Published by under Technology

Human-genome-001In 1990 the Human Genome Project (HGP) was launched, with a goal of sequences all the euchromatic DNA in the human genome within 15 years. The project was completed in 2003, two years ahead of schedule, at a cost of $3 billion. This was one of the great scientific achievements of our species – we set out to complete a huge goal, and we did, similar to going to the moon.

Over the course of the HGP the technology to sequence genes improved by orders of magnitude, becoming faster and cheaper. Not only did we gain a fantastic resource of knowledge that would fuel further science for decades, but the technological advances allowed us to sequence the genomes of many other organisms with still more scientific benefits.

Yesterday scientists announced HGP-Write, a plan to synthesize from scratch an entire human genome. Clearly they hope to replicate the success of the HGP. The plan is to organize an international effort to synthesize a human genome, advancing the technology to do so along the way. The proposed budget is $3 billion, the same as for the HGP.

Continue Reading »

6 responses so far

Jun 02 2016

Theranos Exposed

Published by under Skepticism,Technology

Holmes-Summit-e1460127673162-1200x945I love a good cautionary tale. Perhaps my medical background predisposes me to “post mortem” analysis – what exactly went wrong and why? There is a risk of the cautionary tale, however, in that it is easy to impose a preexisting narrative onto events. The tale can be easily coopted, and then you are not learning a lesson but just reinforcing an existing belief.

I will try to avoid that trap while discussing the story of Theranos, but I am picking this story because it does support a cautionary tale skeptics like to tell.

Theranos was a hot tech startup based on a “disruptive” technology developed by its young maverick founder, Elizabeth Holmes. The story was perfect for the tech industry, who ate it up. It seemed a little dodgy to the medical industry, however, who viewed it with skepticism.

Here is a good overview of the story, with a list of tech news articles fawning over the new startup. Even some news sites who should have known better, like Smithsonian Magazine, were taken in. Read this article – it reads like a marketing brochure for the company. It overstates the limitations and problems with current technology, and overhypes how revolutionary the new technology will be. It asks, but does not answer, the key question – how is this new alleged technology supposed to work.

The claimed breakthrough of Theranos was a streamlined process for laboratory blood analysis that promised to perform 30 tests on a single drop of blood with same-day results. This would eliminate the need for drawing vials of blood and replace that with a simple finger prick.

For any scientist there are immediate red flags. Each blood test, in a way, is its own technology. You don’t measure sodium in the blood the same way you measure glucose, or test for the presence of antibodies to a virus. Yet Theranos claimed to have revolutionized dozens of standard laboratory tests. This would require a massive amount of research and development, or the introduction of an entirely new technology.

Such technology does not come out of nowhere. Research builds upon other research and then is translated into practical applications. The myth of the lone researcher making breakthroughs in their garage is largely just that, a myth. But that image clings tightly to the public consciousness. This makes it easier to sell the narrative of the lone genius making breakthrough technology.

Perhaps the tech industry is especially susceptible to this narrative. A team of coders with a great idea can create a disruptive app that will change the game. Investors are looking for disruptive startups, nerds with a great idea and the next billion dollar company. Medical technology is different, however. There needs to be a paper trail, years of research leading up to the application.

Now that the true story of Theranos is coming out, it seems obvious in retrospect that the whole thing was a scam. First, their labs were conducting 70% of their blood tests on conventional machines using conventional blood draws. They claim this is just while they were waiting for FDA approval of individual tests, but still they were not delivering on their promise.

Second, Theranos just voided the last two years of study results that were being conducted with their technology, what they called the Edison machine. In essence they just admitted that their technology does not work, and the lab test results they provided were not accurate. The company, essentially, has completely evaporated. The company now faces a class action lawsuit.

Last year Forbes estimated Holmes net worth at $4.5 billion. Yesterday they revised their estimate down a bit – to zero.


If this is a cautionary tale, what are the lessons? The obvious one, of course, is to be skeptical. Treat every new exciting claim as if it is a scam until proven otherwise. Don’t buy corporate marketing propaganda. Ask the hard questions, like exactly how does this work?

There are genuine breakthroughs, but most claimed breakthroughs aren’t. The market is currently being flooded with snake oil and medical pseudoscience, so again, it is a good default position that any new claimed medical breakthrough is probably a scam, or at least overhyped.

Also, be skeptical of nice neat narratives. If a story sounds ready made for movie plot, it probably is just that – a fiction. We love stories of underdogs rising from obscurity by challenging the big boys. We love the lone maverick narrative, the young genius, the rugged individual not afraid to break with convention.

The truth is often much less glamorous and exciting. Progress in science is more often made by various teams each contributing their incremental advance. Ideas rarely come out of nowhere. The big advancements are ones that we saw coming 1-2 decades before they became a reality.

Be aware of common red flags: A company selling a new technology should be able to describe, at least in general terms, how the technology works. You can do this without giving away technical secrets. If they refuse to give a satisfying answer for whatever reason, be suspicious.

If the relevant scientific community is skeptical, you should be too.

If they are using science-sounding jargon but do not really make sense (such as talking about frequencies, or quantum mechanics) it is almost certainly a scam.

Theranos is certainly not a rare case. Companies selling pure snake oil are out there in the thousands. Theranos was perhaps just the biggest one. I also doubt that the tech industry has learned its lesson. The allure of billion dollar  disruptive technology is just too great.


17 responses so far

May 26 2016

What’s Killing the Bees

honeybeesYesterday I saw a bumper sticker that stated, “Save the Bees, Buy Organic.” Of course, a bumper sticker is not the place for a nuanced or thorough treatment of a complex topic. It is a venue suitable for simplistic slogans.

People like simple narratives, but reality rarely conforms to our desires. This has led to a frequent reminder, popularized by Ben Goldacre, that you will often find the situation (pretty much whatever situation you consider) is more complex than it might at first seem. That is a good rule of thumb – it is fair to assume as a default that any topic is more complex than your current understanding, or how it is being presented in the media, or how it is understood in the public consciousness.

Complex and ambiguous situations, like the fate of our pollinators, become a convenient Rorschach test for ideology. People tend to impose on this complex and not fully understood situation whatever simplistic narrative suits their beliefs and values, like the notion that organic farming will somehow save the bees. Continue Reading »

14 responses so far

Apr 29 2016

Some Battery Click-Bait

Published by under Technology

battery-24In an updated version of Dante’s Inferno there is one level of science hell that is specifically reserved for headline writers. In their hell there are endless signs pointing the way out of their torment, but all the signs are misleading or overhyped. They are therefore perpetually devastated by the difference between what they are promised and the reality.

Take the following headline (please): Researchers have accidentally made batteries that could last 400 times longer. They know the average person is going to read this and think they won’t have to recharge their cell phones but once per year (it’s recharge day, everybody). The headline is not technically wrong, it is just deliberately ambiguous in the use of the term “last.”

There are two ways in which batteries “last” – how long do they last on one charge, and how many charge-recharge cycles do they have. Guess which one we are talking about here.

The news item is also genuinely interesting, and would have grabbed my attention even without the click-bait headline. But first, a quick primer on batteries.

Continue Reading »

11 responses so far

Mar 21 2016

In Praise of Lab-Grown Meat

Published by under Technology

lab-grown-meatLab-grown meat now seems inevitable, although it is still hard to predict exactly how long it will take to become a popular consumer item. Here is a quick overview of what this is, and its potential to improve our food supply.

Lab grown meat involves taking muscle stem cells from animals, like pigs, chickens, or cows, and then growing them, well, in a lab. They can be grown in a large vat of nutrients.

What you end up with is not fully formed muscle, as if it were taken from an animal, but simply a mass of muscle cells. Animal muscles also contain fat, vessels, and connective tissue, which help give it its texture. For taste the fat marbling is probably the most important.

Continue Reading »

19 responses so far

Mar 14 2016

Patients Prefer Video to Face-to-Face Consultation

atmThere are now many aspects of my life in which I prefer to interact with a computer rather than a person. When I stop to fill up my tank or remove some cash from my bank account, I can do so quickly and efficiently by interfacing directly with an automated machine. It is interesting to think about exactly why this is, but first let me discuss the findings of a new study looking at this very question.

Dr. Matthew Winter and his fellow researchers presented the results of a randomized trial at the European Association of Urology 31st Annual Congress in which they compared giving informed consent via video vs in person prior to a urological procedure.

The study randomized 88 patients to either have a face-to-face consultation with the surgeon, or to viewing a prepared video including animation to describe the procedure. They then quizzed them on their knowledge of the procedure, and did a follow up cross-over in which the groups switched and then were re-tested.

Continue Reading »

5 responses so far

Next »