st-musings

[st-musings]

New laws of robotics: Defending human expertise in the age of AI, Frank Pasquale

=============================

New laws of robotics. 

Robotic systems and AI should complement professionals, not replace them 

Robotic systems and AI should not counterfeit humanity 

Robotic systems and AI should not intensify zero sum arms race 

Robotic systems and AI must always indicate the identity of their creator(s), controller(s), and owner(s)

Clever marketing made some mental health apps as a cheap convenient alternative to therapist. 

Pragmatic modernizers emphasizes the vocational training for jobs.

Behaviorists offer drilling methods to rapidly inculcate knowledge. 

Top down prescriptions of excellence are now common among policy elites. Their rationalization of education as workforce development fuses vocational and behavioural tendencies. 

To the extent leaders want workers of the future to be bossed around by robots or algorithms,  it would make sense to rout them into schools with mechanistic lesson plans and highly controlled teachers, culminating in robotization of the educational workforce. 

Pressey 1924 automatic teacher. Pressey testing machine let students choose one of five options and offered immediate feedback. Each press of a button would advance a mimeographed sheet further revealing the next question. Skinner believed the teacher is a mere reinforcing mechanism. His teaching machine let students pull levers to record their answers and turned on a light when a correct answer was registered and even dispense a pellet of candy once a student had answered enough questions correctly. 

Audrey Watters compiled an encyclopedic array of edtech dreams - virtual reality classrooms, matrix l-like instant learning through brain stimulation 

When financial inclusion becomes creepy and predatory. 

We may want our doctors to access such information, but we need not let banks, employers, or others use it. There is more to life-and public policy-than consenting transactions among borrowers and fin-techs to ensure "financial inclusion"

Without proper guardrails, there will be a race to the bottom in both data sharing and behavior shaping as more individuals compete for better deals. That would result in a boom in predatory in-clusion, creepy inclusion, and subordinating inclusion.

Lawmakers should discourage or ban each of these types of "inclusion."

Following philosopher Michael Walzer, we might categorize the family and many civil-society institutions as distinctive "spheres of justice"-realms of human experience that ideally operate according to their own logic of distribution, rather than standards externally imposed.

But one core idea of Walzer's is that one's standing in one sphere should not be unduly affected by one's conduct in another. In other words, simply because a person goes bankrupt or speeds, that should not seriously reduce his reputation as an employee, minister, or parent, much less that of his relatives.

Why should a person be charged a higher rate of interest merely because she bought a brand of beer frequently purchased by defaulters? In the realm of black box AI, this makes perfect sense-what matters is prediction and correlation, not justice or fairness. 

If, however, we consider credit a sphere of justice-something implicit in all the moral discourse surrounding obligation, bankruptcy, and promises-the correlationist impulse makes little sense. A person deserves to be judged on her own merits, in a way that is publicly intelligible and legitimate. 

There are distinctive patterns of human conduct that can be performed better or worse, more or less skillfully, but whose quality is better described and prescribed in a nar-rative, holistic, and emotive mode rather than reduced to numbers or coded into instructions. 

The microeconomic perspective of individual businesses is to minimize labor costs. The macroeconomic perspective of society as a whole requires consumers to have some source of income-and for most of us, that is work. How society balances these rival demands of labor and capital is a subject of both politics and economics-also known as political economy. A political economy perspective opens possibilities that are now dismissed as impractical, or not even considered at all.

Standard accounts of AI economics tend to pose a trade-off between regulation and innovation. Political economy rejects that idea.

We are not trying to maximize the size of some imaginary economic "pie" and then redistribute the resulting bounty. Rather, law can and should shape the type of Al we get. 

Widespread diffusion of knowledge and other capabilities needed to govern and improve technology deployment. 

This is less an economic principle than an adaptation of an ideal of governance known as subsidiarity, this principle commends a devolution of responsibility to the most local entity capable of handling it well. 

Some captains of computationally driven industries may balk at this suggestion. Anthony Levandowski, an engineer behind some key advances in self-driving cars, once commented that "the only thing that matters is the future. I don't even know why we study history. It's entertaining, I guess-the dinosaurs and the Neanderthals and the Industrial Revolution, and stuff like that. But what already happened doesn't really matter. You don't need to know that history to build on what they made. In technology, all that matters is tomorrow." This dismissal of the past is troubling. Those at the commanding heights of transport technology need to understand why, for example, car culture may have been an enormous mistake. A deeply learned and humane book like Ben Green's The Smart Enough City tells that story with authority and should be read by anyone rushing to double down on automobile-bound individualism when collective transport solutions are so critical to reducing greenhouse gases. 

At present, economists and engineers dominate public debate on the "rise of the robots." The question of whether any given job should be done by a robot is modeled as a relatively simple cost-benefit analysis. If the robot can perform a task more cheaply than a worker, substitute it in. This microeconomic approach to filling jobs prioritizes capital accumulation over the cultivation and development of democratically governed communities of expertise and practice. As AI and robotics both play a larger role in services and suffer from serious problems like bias and inaccuracy, we need these communities of expertise ever more urgently. We do not simply need more AI, we need better AI, and more technology designed to augment workers' skills and income. To do so we must move beyond the biased frame of the "cost disease" to the "cost cure".

[st-musings]

From breakthrough to blockbuster, the business of biotechnology Donald Drakeman

============================

Recombinant DNA, monoclonal antibody twin foundations of modern biotechnology

Antisense technology, gene therapy, tumour vaccines, stem cell therapy, combinatorial chemistry, high throughput screening, gene chips, tissue engineering, bioinformatics,proteomics, rational drug design, novel delivery technologies

Tufts study

Estimated cost of drug development at $2.56 billion

After all these opportunity costs are factored in the actual money spent in the drug development process is estimated at $1.4 billion

Only a tiny number of the largest biotech companies have integrated R&D organizations capable of discovering new product candidates and then developing them all the way to the commercial market.

Contract research organizations enabled "virtual" biotech companies. The expertise of CROs spans the entire spectrum of the drug development process, from creating initial compounds to performing in vitro, in vivo and human clinical testing to manufacturing the necessary quantities of the product to applying for FDA approval and even providing contracted sales and marketing services.

The drug development process

Academic research - many thousands of ideas

Early research and preclinical - invitro invivo testing

IND submission (Investigative new drug)

Clinical trials - phase I safety, phase 2 safety and efficacy, phase 3 safety and efficacy at large scale

NDA/ BLA submission

FDA approval

In looking at drugs entering clinical development 11.83% of the product candidates had reached FDA approval

Bayh-Dole Act university patent technology licensing bolstered economy by$1.3 trillion. Life sciences accounted for 70% of licenses and 93% of gross technology transfer revenues.

Few cases where venture capitalists wait to successfully develop a new drug. They are focused on having the biotech company

achieve whatever technical and corporate milestones will create opportunities for a successful exit

Since there are dramatic ebbs and flows in the overall availability of investment capital for biotech companies, there can be a boom-or-burst feeling in the early stage biotech arena, irrespective of the rate at which exciting new technologies and products emerge from research universities and other medical centers.

Qualities for biotech entrepreneurship.

Do you always think there is a better way to do things?

Are you willing to take on just about anything, even if you don't know much about it?

Are you comfortable taking risks?

Do you like to do new things, or do you prefer routine?

Can you accept rejection and failure?

Why Biotech companies are more innovative than pharmaceutical companies

In contrast to a large centralized environment that can be prone to limiting the overall number of projects and then be slow to stop the unsuccessful one,a decentralized environment of multiple external investors maximizes the potential for following the two critical principles of (1) initiating many diverse projects and (2) stopping the ones that are not working out. Having many different decision makers who are responsible for allocating funding creates a favorable environment for trying many different things. It also minimizes the effects of the sunk cost fallacy and the intra-organizational perspectives that make it difficult for large, cemtralized structures to make responsive termination decisions.

In fields outside the life sciences, technological advances often lead of ways to do things faster, better amd cheaper.

The crucial financial point is that biotech's breakthroughs may be lifesaving but rarely been cost saving

Scientists and physicians can figure out if a new drug actually extends lives, and mathematicians can calculate the costs; but none of those analyses lead directly to a considered judgment about who should have those benefits and at what price.

National Institute of health social value judgments 4 principles

respect for autonomy

non-maleficence

beneficence

distributive justice

Eg monoclonal antibody technology was discoved in England but is so expensive that the NHS refused to pay for them. It will be cold comfort to know that the UK economy was stimulated by research funding that contributed to the development of a new drug if that stimulus was not financially potent enough to allow the nation to be able to afford to pay for the drug itself

[st-musings]

Smart cities big data civic hackers and the quest for a new utopia, Anthony Townsend

============================

Trade offs of smart cities: efficient but preserve opportunities for spontaneity and sociability. Secure but not surveillance chambers. Open and participatory but provide support structure for those who lack the resources to self organize. Need to be inclusive. "The death and life of great American cities" Jane Jacobs - cities have the capability of providing something for everybody when they are created by everybody.

The unintended consequences of nee technologies dwarf their intended dedign. Motorization promised to save city dwellers from factory smoke but scarred the countryside with sprawl.

Cybernetics. Rio de Janeiro city control room 2012 - remote control city. Norbert Weiner.Jay Forrester - computer models for urban planning.

RAND New York fire department simulation. Oversimplification of model to suit limitation of the computer. Portland 2011 IBM system dynamics for smarter cities. Michael Batty Centre for advanced spatial analysis. Models assumed a closed loop moving from one equilibrium to another and no effect by the external environment.

Patrick Geddes - civic participation

Ebenezer Howard solution for urbsn renewal - start over. Price of city planning is cities built for cars. Tyler Cowen no more big breakthroughs "great stagnation". Hal Varian, rapid transformation.

The history of the internet. TCP/IP favoured by academics won over X.25 backed by telecommunications companies. Cisco, IBM, Siemens planning for smart cities is to cobble together internet of things, analytics, video communications no bottom up everyday people participation. Technology giants designs are 21st century upgrade to 20th century paternalism these design fail to realize the full potential of smart cities. The bureaucrat leaves out time, patience, detail, inter-relating past and future, human purpose, citizens above mechanical requirements.

Dodgeball- crowdsourced online city guides.

Internet of things- not only a system for remote monitoring but also a platform for citizen microcontrol of the physical world. Arduino gives us the tools to structure intelligence into the human scale objects we live in. A social technology that get people talking to each other. Cheap microcontrollers Arduino boards at $25 are networked into clusters.

Foursquare- local, social, mobile software

In 2007 hackers figured how to jailbreak the iPhone. In 2008 Apple co-opted the movement by launching the App store

We can ship laptops to the world's slums but can force anyone to use them. Map Kibera demonstrates how open source tools can empower them to create knowledge relevant to the problems they face. The advent of the digital age means that gatekeepers to information can be bypassed allowing for a new parallel information system to be created by marginalized citizens. Aa fashionable crowdsourcing has become in the developed world it is regressive. Governments may withdraw services as citizen-driven alternatives expand and allow governments to free themselves from the obligation equalize services for the poor.

Apps contests on open city data was a model that could deliver innovation wth nearly zero funding. Problem with apps contests driven by government data is they rely on programmers to define problems instead of citizens or government.

Problem of parochialism in government software procurement. Every city wants to support the development of a local technology industry. Reinvent the wheel and dress it up as an innovation program. Variations of the same service, subpar apps and different one for each town.

"Connected cities" Haselmayer

Contrast to situated software

Rebound consumption or Jevons paradox: improve roads or have driverless cars, same congestion as people buy more cars. Better power efficiency, costs lower, people buy new appliances

[st-musings]

The entrepreneurial state debunking public vs private sector myths

=========================

The entrepreneurial state debunking public vs private sector myths, Mariana Mazzucato

Financialization of private sector spending less on R&D large companies share buyback government and small companies are investing in R&D. Socialization of risk privatization of rewards

Karl Polanyi: it is the state that imposed the conditions that alloed a market driven economy. Capitalism thought of as market driven has been shaped by the state since day 1

Rise in patents does not reflect a rise in innovation. But rather patents used for cross licencing, wider range of types of patents, attract financing. Rise in patents most of them with little worth

The notion that the green revolution can be led by the business sector with a nudge by the state is wrong. China makes available 47 times more money that solar companies can use Britain is fiddling with 'play money'. If VC aren't interested in capital-intensive industries or in building factories what they offer in economic development is limited. Public finance through development banks is superior to VC or commercial banking in fostering innovation because it is committed and patient

Bank bonuses should not be criticized using arguments against greed rather that such compensation is a reflection of risks taken

When the appropriation of rewards outstrips the bearing of risk in the innovation process, the result is inequity; when the extent of inequity disrupts investment in the innovation process, the result is instability; when the extent of instability increases the uncertainty of the innovation process the result is slowdown in economic growth

Policy recommendations: golden shares of intellectual property rights and a national innovation fund, income-contingency loans and equity and development banks

[st-musings]

Get outdoors. Light up your life

[st-musings]

Debunking the baby development myths

========================

There is a myth that babies who listen to Mozart before or as infants gown up to be geniuses. This Mozart Effect has spawned a multi million baby edutainment industry. The myth originated from an 1993 study in Nature by Rauscher, Shaw and Ky that college students who listened to a Mozart piano concerto before taking a test that measured visual-spatial abilities did better than those who received relaxation instructions or silence. But in fact all the music does is to heighten alertness and the effect dissipated after 10 minutes.

So in the end, money would be better spent on on music lessons for kids and improves their ability to concentrate; helps their self-confidence; helps their coordination; among other positive developmental benefits to the brain and these effects seem to last for years or their entire life. Children are natural learners and each child is better off learning by curiosity and we develop resilience, risk-taking and self-regulation. In fact on the contrary children learn less from videos that from playing actively.

https://www.todayonline.com/commentary/uncovering-myths-early-children-education http://www.kathleenhowland.com/blog/2016/10/31/the-effect-of-the-mozart-effect http://www.todayifoundout.com/index.php/2010/10/listening-to-mozart-wont-make-you-smarter/

[st-musings]

This season Arsenal's Ah Lau means business.

[st-musings]

Singapore 1 Chinese Taipei 2 Perhaps we may reflect on Singapore’s football as a microcosm of society at large. In 1973 I watched us beat Vietnam in the SEAP games, in 1994 I travelled to Shah Alam to watch us lift our last the Malaysia cup. That was to pave the way for the S-league to develop a wider pool of players. Then in the aftermath of the 1998 world cup GCT saw the success of the black players in the France team and we pushed for the foreign talent scheme.

Fast forward today, we at ranking 165 seemed to have stayed stagnant or grown negligibly, while our neighbours have improved and overtaken us. We have been overtaken by Philippines, Laos. In Asean we can only beat Cambodia. At this rate soon we can only win against Timor Leste.

In the larger society we may have been distracted by the hubris of past success, attempted to ride the same formula for future success and let slip our neighbours who have no baggage of the past.

[st-musings]

Terms of Service. Jacob Silverman ========================= Digital technology besides transforming, improving lives, liberating also creates surveillance. We cannot assume passivity and inevitability. But scepticism and criticism are necessary to counter surveillance and protect privacy against the internet giants. Ray Kurzweil, Google, the Singularity, self aware machines upgrade brains to the cloud Katherine Losse , The boy kings, Companies over countries, speech for Zuckerberg Likes, taggable data allows emotions, every online action to be mined more opportunities to collect user information Virality isn’t about apportioning game or anointing the day’s popular meme. It is the larger system of quantification, surveillance and feedback loops that underlines social media. While the he social media era allows us to read far more broadly and more journalists can reach a wider audience, it also leads to churnalism cheap disposable content, seeking faster to publish, more followers, ephemera by volume instead of depth. The best paid people in a news site are the data analytics team who look at what’s trending and hand it to the editors to craft the most searchable headlines. Writers are lowest end poorly paid or unpaid aspiring journalists with promise of future fame. The sharing economy is just a way for companies to outsource labour and Risk to individual consumers who take on the traditional responsibilities of a company while getting a fraction of the services. TaskRabbit, Lyft, Uber, Sidecar are just aggregators calling themselves software companies not employers or transportation companies, contributing little, outsourcing risk and pocketing the most profits. There is the hidden costs of unacknowledged work, car cleaning, extra liability of driver, dangerous misrepresented tasks. In the process depriving local government of revenue from taxes. The sharing economy isn’t some fairer, communitarian system of exchange but an extreme form of capitalism in which everyone is an entrepreneur but no one is employed. Everyone works but for a beggar’s bargain. Dave Eggar The Circle The knowledge economy relies on extracting the maximum information and data from users at minimum cost. Facial recognition digital billboards in future will individualize ad targeting. Ads follow us around the internet throughout the world. It’s surveillance, stalking, harassment, visual pollution. Problems. Reputation, viral fame, micro-labor, page view fueled journalism, intrusiveness of digital advertising, unannounced data collection.

[st-musings]

China bike sharing phenomenon Everywhere in the big cities you can rent a bike for prices as low as 7 euro cents (½ RMB) per half an hour. How do the bike sharing companies make money? Maybe not in the bike rentals per se. From the venture capital? From government investment? From the deposits of the millions of accounts? In SG $49 for Obike x all the users! Ofo, Chinese bike-sharing start-up, says it's now worth more than $2 billion! Didi Chuxing-backed ofo, founded in 2014, started its business by providing university students and teachers bike rental services inside university campuses then expanded into cities with financing from big name investors including the America-based Coatue, Xiaomi, and CITIC Private Equity Funds Management Co.  Mobike, founded in Shanghai in 2015, has more advanced bikes, with built-in GPS trackers. Its wide array of investors include tech giant Tencent, electronics supplier Foxconn, and Singapore-based Temasek But for the consumer it's good news. Cheap bikes, minimum investment. Enjoy the ride.

[st-musings]

The Twisted Perversion of Racial Diversity in Singapore

- Libertarian Society Singapore

http://www.libertarianism.sg/2016/10/The-Twisted-Perversion-Racial-Diversity.html?m=1

[st-musings]

Committee for future economy report

The CFE has released it’s report on 9 Feb17 with these strategies https://www.gov.sg/microsites/future-economy/the-cfe-report/7-strategies 1. Deepen and diversify international connections 2. Acquire and utilise deep skills 3. Strengthen enterprise capabilities to innovate and scale up 4. Build strong digital capabilities 5. Develop a vibrant and connected city of opportunity 6. Develop and implement Industry Transformation Maps (ITMs) 7. Partner each other to enable innovation and growth

[st-musings]

Antibiotics harm our microbiome

Our gut harbors millions  of good micro-organisms and probiotics replenish them, when their balance gets disturbed. They boost our immunity as our body teams up with these good microbes.

[st-musings]

What is Kefir and how it benefit our health

Kefir is a fermented milk that improves your general health and well-being by first improving your digestive health and boosting your immune system,  Containing microorganisms called probiotics it cleanses your intestinal tract and kill bad bacteria. Kefir is safe for children, the elderly, the lactose intolerant, diabetic patients and even pregnant women.