Blog

The Evolution of Test

“Everything is a test,” says Terry Pratchett in "I Shall Wear Midnight".

Test - a way to establish if something is acceptable or not - is one of the world's most commonly performed procedures. Billions of years ago, Mother Nature established agile processes to test molecules, cells and organisms while creating them. Stone age toolmakers were checking if their tool was doing what it was supposed to do before using it. Ancient humans assessed others for civil service and education. How did all those tests evolved with time?

First tests were manual. Teachers read students' essays, engineers manually debugged their code, and a bridge was tested by walking an elephant across its length.

The need to scale up led to automation. Educators got standardized tests and scanners. Software Quality Assurance professionals mastered machine-driven test cases and new test automation frameworks, teaming with software developers for better productivity. Medical testing evolved from tasting urine to sophisticated techniques and molecular diagnostics. At the forefront of test automation, electronic design engineers moved from asking "Does it work?" to "Are all elements present and working?" to "What could go wrong with this design?" focusing on defect-, circuit-, environment- and equipment-dependent variations.

What's next?

Will artificial Intelligence take over, testing software, hardware and interviewing people for the AI-proof jobs? Will the Internet of Things and Wearables improve the assessment of people's health, educational progress and behavior? Will medical diagnostics merge with therapeutics enhancing nature's proof-reading and error correction mechanisms? Will crowdsourcing evolve into testsourcing with everyone everywhere being a tester of something for public good?

Perhaps. We have already started to explore AI-powered bots testing apps and see Artificial Intelligence challenging medical doctors on their home turf, so anything is possible.

0 Comments

Do You Want AI with that?

4 Comments

When you hear about Artificial Intelligence (or AI, in cyber parlance), you may picture Ex Machina, Azimov's robots, Hal 2001, R2D2, Terminator, Siri, Alexa, Cortana, Watson or Google Brain.

You may also recall flashy news headlines about AI predicting specific health events and outcomes more accurately than a doctor, navigating better than humans, outperforming government workers, bankers, trial jurors and psychologists.

But there are many other new ideas out there.

This April, Y combinator's online startup school opened its doors to about 3,000 companies from 141 countries. 169 of them indicated "Artificial intelligence" as their industry and 62 of them "graduated" with a video-pitch, although many more mentioned AI technologies in one way or another. What are young startups working on?

Digital assistants and web robots (aka bots) are eating the world. Millennials are willing to rely on chatbots. Two out of five trust bots more than humans (from 30% in health-related information to 74% in regards to food). Most would prefer chatbots for communication compared to texting and email, phone calls and face-to-face interaction. The total number of messaging and voice chatbot startups in the class of 2017 was close to 50.

The new generation is thinking ahead creating chatbot factories, nurseries, marketplaces - where bots can be auctioned off - and schools - where you can feel assured your bot is cared for. Schools for humans are also proposed - to learn how to better understand your bots.

By 2020 30% of mobile/web browsing sessions will be done without a screen. Many startup school founders are looking at AI as an interface - to buy and sell (fashion items, stocks trading and sports betting, real estate, whatever), get help with mundane work, learn new languages, improve our personal style, save the planet, read news, stay energized, eat well and have a better sleep.

One stealth startup even wanted to convert anything we touch to artificial intelligence, Asgard proposed to use AI technology to discover technologies, and Nudge to motivate learning artificial intelligence - getting closer to the 3rd generation AI.

Perhaps the next step is to create artificial intelligence tools to develop the next generation artificial intelligence tools.

The future is AI, but we are still competitive.

4 Comments

Let those who never smelled bad cast the first stone

6 Comments

Analysis of our metabolism is crucial to comprehending the responses of our genes and microbes to the stresses of daily life, and to elucidating the causes and consequences of health and disease.

And measurement of urinary metabolites - small molecules produced from foods, drinks, drugs, environmental contaminants, bodily waste products and bacterial by-products -
is key to the analysis.

We applied this approach to an elusive condition that has always evaded diagnosis: socially and psychologically distressing odors that occur without a known or apparent cause.

The figure below summarizes preliminary results of metabolic profiling performed by Dr. Wishart's research group, using direct Flow Injection Mass Spectrometry (DI-MS).

The first part of our study explored 95 metabolites: biogenic amines (odorous by-products of microbes predicting certain health conditions), amino acids (for diagnostics of endocrine disorders, liver diseases, muscle diseases, neoplastic diseases, neurological disorders, nutritional disturbances, renal failure, and burns), acylcarnitines (biomarkers of fatty acid oxidation disorders and several organic acidurias), phospholipids and sphingolipids (useful in evaluating some enzyme deficiencies) in samples of 15 participants.

No measurement differed from normal population values - individuals that never complained of any unusual odor problems or other notable medical issues) - even for the "smelliest" compounds such as Putrescine (rotting fish smell) or Spermidine (sperm-like or bleach-like smell). Moreover, average values for these metabolites were actually lower in our cohort, as were values for selected aminoacids, although some aminoacids, biogenic amines and lipids were slightly higher than averages for normal population. So there were definitely significant albeit subtle differences in metabolism between the control and test group, as well as among the participants.

We combined results for all 95 metabolites using a uniform standard scale. A statistical technique, Principal Component Analysis (PCA), uncovered combinations that best differentiated the participants. The figure shows that there are at least two or three different types of profiles. Is there something in common between participants in each group?

Our questionnaire shows that there is, and it is not the diet!

The most significant differences between participants clustered into the groups shown in the figure are their environmental sensitivities (hence their metabolism) and their symptoms.

Participants from the first group (green dots) had very different diets ranging from plant-based, low choline, low carb and Paleo diets to standard American and junk food diets. They seemed to be more tolerant to sugar than other groups, but more sensitive to dairy/lactose. All participants reported exhibiting sour odors including vinegar, ammonia, fishy and urine-like odors. When the usual odor was less sour (but nonetheless unpleasant), it became more sour with lower blood sugar (for example, with alpha lipoic acid supplements). Chlorophyll supplements led to a grass- or weed-like smell. This group was similar to the "sour" group from our earlier study in UK (NCT02692495, discussed in this video).

The second group (purple dots) was similar to the "sweet" group from the earlier UK study. Participants seemed to be more sensitive to sugar and carbohydrates, and to environmental triggers. Main odors experienced ranged from fecal, sewage, cheese-like, rubber, burning, tobacco smoke (participants were non-smoking), barn, rotten eggs and mold. Dairy could make odors "cheesier".

The third group was also sensitive to sugar, but the main difference with first and second groups was structural gastrointestinal issues. This group had a different diet from other groups too, it was 98-100% vegan. But further testing is needed before any conclusions can be drawn.

We observed a few other interesting patterns but will wait before commenting on them until more data is collected.

Meanwhile, if you have never experienced any unpleasant odors in your life, or experienced some that did not affect your social life, you can contribute to our analysis of the human metabolism by participating in this anonymous survey.

References

Irene S. Gabashvili (2017) Community-led research discovers links between elusive symptoms and clinical tests. Submitted

Bouatra, S., Aziat, F., Mandal, R., Guo, A., Wilson, M., Knox, C., Bjorndahl, T., Krishnamurthy, R., Saleem, F., Liu, P., Dame, Z., Poelzer, J., Huynh, J., Yallou, F., Psychogios, N., Dong, E., Bogumil, R., Roehring, C., & Wishart, D. (2013). The Human Urine Metabolome PLoS ONE, 8 (9) DOI: 10.1371/journal.pone.0073076

Irene S. Gabashvili (2017). Raw Data from Clinical Trial NCT02692495 Mendeley Data, 1 DOI: 10.17632/8bk6h6bmkr.1

6 Comments

Sharing the Future with Artificial Intelligence

2 Comments

Artificial intelligence has reached a buzzword utopia as it seems everyone is talking about self-driving cars, delivery drones and virtual assistants with human-like "intelligence." Some believe this new era of AI will make the American Dream universally accessible, enabling early retirement in bucolic settings. Others are concerned about a greater inequality created by a jobless future.

Technological advancements have connected billions of people (and digitally divided the rest), merged real and virtual worlds and enabled our collaboration with robots - at least in the form of software that streamlines everyday tasks.

Yet, since the beginning of the third industrial revolution about 40 years ago, job insecurity has steadily increased as the life cycle of relevant skills has decreased (Blanchflower and Oswald, 2011).

This trend could continue with the AI-enabled fourth industrial revolution. While having the potential to significantly improve productivity, AI could dramatically reduce the need for workers and present serious challenges for healthcare industries, governments and individuals (Schwab, 2015).

One option for reducing financial anxiety related to the decreased demand for human labor is to provide subsidies to every adult. A recent US survey found that 46% of people support the idea of a universal basic income, formulated as a "negative income tax" (Milton Friedman), guaranteed income (Martin Luther King Jr.) or annual cash grant (Charles Murray). The idea did not pass the vote in Switzerland, but is under consideration in India, Finland and Oakland, California. Earlier case studies - such as the Alaska Permanent Fund, established 40 years ago, has revealed unintended consequences: the program encouraged retirees to move to Alaska, squeezed wages, fostered addiction to "free money", and created a “consumption-frenzy” at local shops when the checks were distributed. Last month Alaska House voted to lower the dividend, and this could hit middle-income residents relying on it.

But even if the government provided a stable money supply, would it be easy to find a purpose in the jobless future?

Gig economy, volunteerism and citizen science could potentially provide such a purpose. But Uber-like platforms are built for "lone wolf" and low skilled kind of jobs, while collaborative research, in its current form, works best for environmental sciences. Besides, it increases expectations that research for the public good should be done by volunteers rather than paid professionals.

To prepare for the new automated future, we need completely new collaboration platforms that help master and utilize high level skills. We'll share our thoughts about it next time.

REFERENCES

David G. Blanchflower, & Andrew J. Oswald (2011).
International Happiness: A New View on the Measure of Performance Academy of Management Perspectives, 25 (1), 6-22

Klaus Schwab. The Fourth Industrial Revolution. World Economic Forum; 2015

Goldsmith, Oliver Scott. "The Alaska Permanent Fund Dividend: A Case Study in Implementation of a Basic Income Guarantee." (2010).

Painter, A. (2016). A universal basic income: the answer to poverty, insecurity, and health inequality? BMJ DOI: 10.1136/bmj.i6473

Forget EL. The town with no poverty: using health administration data to revisit outcomes of a Canadian guaranteed annual income field experiment. 2011.

Brougham, D., & Haar, J. (2017). Smart Technology, Artificial Intelligence, Robotics, and Algorithms (STARA): Employees’ perceptions of our future workplace Journal of Management & Organization, 1-19 DOI: 10.1017/jmo.2016.55

Pfeiffer S (2017). The Vision of "Industrie 4.0" in the Making-a Case of Future Told, Tamed, and Traded. Nanoethics, 11 (1), 107-121 PMID: 28435474

2 Comments

More than Meets the Eye

1 Comment

Eyeglasses are almost as old as the civilization itself. They have not changed much since Benjamin Franklin's bifocals in the 18th century. Nor were they made obsolete by laser surgery and contacts. Still, eyeglass technology leaves much to be desired.

There has been a lot of research to improve the limitations of bifocal lenses and reduce the side effects - headaches, dizziness, muscle fatigue and the small field of view. Introduction of trifocals and progressive multifocal lenses did not solve all the problems.

New materials and technologies may help to interactively adjust the lens in glasses or change focus on the fly.

Like self-adjustable glasses invented by Josh Silver (shown on the right). Fluid-filled lenses promise to correct a wide range of vision problems, including nearsightedness, farsightedness and inability to focus. No need to see an optometrist that most people in the developing world can't afford.The price of these glasses, however, remain too high even for the developed world - as the quality is still inferior to inexpensive off-the-shelf glasses that can be easily test-driven in the drug store.

One of the new lens technologies gaining most attention is adjustable-focus eyeglasses.

8 years ago, adjustable-focus Superfocus' glasses with fluid-filled lenses, were introduced to the commercial market. They glasses required manual adjustment of the focus - by moving the slider on the nose bridge, looked "clunky and ugly", were priced at $900 (later discounted to about $600), but grew a loyal following. Nevertheless, the company went out of business, and several new companies since kept trying to step into the gap. Adlens and Dial Vision are now offering glasses adjustable for both near or far-sighted for less than $20 (shown on the right).

The early prototype of new glasses automatically focusing on whatever the wearer is looking at was displayed at this year's Consumer Electronics Show. The glasses feature tunable lenses from liquid glycerin. An infrared sensor in the bridge of the glasses calculates the distance between the glasses and the closest objects and controls the shape of the lenses. It takes only 14 milliseconds to re-adjust the focus. But the glasses resemble professor Fransworth's eyewear (the mad scientist from Futurama, pictured on the left).

The inventors hope that after a few technological and styling improvements, the product will hit store shelves in two or three years.

Science fiction stories have ambitious vision for the future eyeglasses. Geordi La Forge's VISOR furnished information - such as images or temperatures - directly to his brain. Marty McFly Jr. used his glasses to watch shows. In Surrogates, glasses enable to watch one's own life - as it is lived by surrogate robots. Nada's glasses identify aliens. Harry Tasker's glasses help to have a 360 all-around-vision.

Some of these technologies are slowly becoming mainstream. Google glass has showcased the potential of smart eyewear. Microsoft's patent suggested to use glasses to detect other people's emotions. AVG glasses use infrared light to recognize faces. Dispelix brings visual information directly into the user's field of vision, to prevent eye strain. X-ray glasses from Even Medical enable to see veins. BMW is developing glasses to help with driving. Virtual reality is still boutique technology, but is gaining steam. A recent paper showed that Augmented Reality - when an Unmanned Aerial Vehicle operator can see both the mission plan as well as the video stream captured by the drone - significantly improved situational awareness.

New and exciting technologies are unfolding before our eyes. So, perhaps, smart glasses will come back smarter - letting us see the world like we've never seen before.

REFERENCES

Douali MG, & Silver JD (2004). Self-optimised vision correction with adaptive spectacle lenses in developing countries. Ophthalmic & physiological optics : the journal of the British College of Ophthalmic Opticians (Optometrists), 24 (3), 234-41 PMID: 15130172

Gudlavalleti VS, Allagh KP, & Gudlavalleti AS (2014). Self-adjustable glasses in the developing world. Clinical ophthalmology (Auckland, N.Z.), 8, 405-13 PMID: 24570581

Hasan N, Banerjee A, Kim H, & Mastrangelo CH (2017). Tunable-focus lens for adaptive eyeglasses. Optics express, 25 (2), 1221-1233 PMID: 28158006

Ruano S, Cuevas C, Gallego G, & García N (2017). Augmented Reality Tool for the Situational Awareness Improvement of UAV Operators. Sensors (Basel, Switzerland), 17 (2) PMID: 28178189

1 Comment

<<Previous

Forward>>

AURAMETRIX

The Evolution of Test

Do You Want AI with that?

Let those who never smelled bad cast the first stone

References

Sharing the Future with Artificial Intelligence

More than Meets the Eye

Archives

Categories

Environment
Irritable Bowel
Olfactics
Technologies

AURAMETRIX

The Evolution of Test

Do You Want AI with that?

Let those who never smelled bad cast the first stone

References

Sharing the Future with Artificial Intelligence

More than Meets the Eye

Archives

Categories

EnvironmentIrritable BowelOlfacticsTechnologies

Environment
Irritable Bowel
Olfactics
Technologies