Computer vision: Why it’s hard to compare AI and human perception
In the seemingly endless quest to reconstruct human perception, the field that
has become known as computer vision, deep learning has so far yielded the most
favorable results. Convolutional neural networks (CNN), an architecture often
used in computer vision deep learning algorithms, are accomplishing tasks that
were extremely difficult with traditional software. However, comparing neural
networks to the human perception remains a challenge. And this is partly because
we still have a lot to learn about the human vision system and the human brain
in general. The complex workings of deep learning systems also compound the
problem. Deep neural networks work in very complicated ways that often confound
their own creators. In recent years, a body of research has tried to evaluate
the inner workings of neural networks and their robustness in handling
real-world situations. ... The researchers note that the human visual system is
naturally pre-trained on large amounts of abstract visual reasoning tasks. This
makes it unfair to test the deep learning model on a low-data regime, and it is
almost impossible to draw solid conclusions about differences in the internal
information processing of humans and AI.
How To Close The Distance On Remote Work: The Most Important Leadership Skill
In terms of mindset, your perspective is important. One of my colleagues (an
especially responsive leader herself) says her grandmother has a gift for
making each grandchild feel valued and unique. Great leadership is like this
as well. While no one should play favorites, it’s powerful for each team
member to feel they matter and know you appreciate them and their
contribution. When you give people responsibility and trust them to do good
work, you won’t have to be as involved in the work they’re doing. Your time
will be spent coaching, developing and making decisions where your perspective
or position are most critical. You should set guardrails—for example spending
more than a certain amount of money or which key topics require your input or
decision-making—but within those boundaries, set people free. By not being too
deeply in the details, you’ll have more time to be accessible where you’re
needed most. Another mindset to help you be more responsive is to know
your people well. When you have a good sense of what motivates each employee
and what their unique needs are, you’re able to tune your messages. You’ll be
more responsive when you’re able to meet employees where they are and provide
the information or direction they need most.
2035's Biggest AI Threat Is Already Here
Unlike a robot siege that might damage property, the harm caused by these deep
fakes was the erosion of trust in people and society itself. The threat of
A.I. may seem to be forever stuck in the future — after all, how can A.I. harm
us when my Alexa can't even correctly give a weather report? — but Shane
Johnson, Director of the Dawes Centre for Future Crimes at UCL which funded
the study, explains that these threats will only continue to grow in
sophistication and entanglement with our daily lives. "We live in an
ever-changing world which creates new opportunities - good and bad," Johnson
warns. "As such, it is imperative that we anticipate future crime threats so
that policymakers and other stakeholders with the competency to act can do so
before new 'crime harvests' occur." While the authors concede that the
judgments made in this study are inherently speculative in nature and
influenced by our current political and technical landscape, they argue that
the future of these technologies cannot be removed for those environments
either. HOW DID THEY DO IT — In order to make these futuristic judgments, the
researchers gathered a team of 14 academics in related fields, seven experts
from the private sector, and 10 experts from the public sector.
Fintech 2020: 5 trends shaping the future of the industry
One thing a consumer prefers the most would be, multiple services across one
platform. Many Fintech brands have already rolled out this process of offering
multiple services across one app, but the increase in offerings of robust
solutions through powerful API integrations will add on. In the coming days,
consumers who need banking services are likely to turn to those financial
players, who can offer convenience and ease of transactions that is entirely
safe and secure. To address these consumer needs, banks cannot do much, but
technology can help a lot in digitalizing consumer demand. Blockchain and Big
Data are two technologies in full swing, but they are also two complementary
technologies. According to experts, brands adopting burgeoning blockchain
technology will benefit the most. Financial services will be able to reduce
fraudulent activities, phishing attacks and ensure secure payments. One of the
other things that Fintech needs to bring their attention to is—Artificial
Intelligence, Machine Learning and Data Analytics. As all these can help
financial services in addressing their key challenges like cost reduction and
scrutinize risky transactions.
The dark side of Israeli cybersecurity firms
The common denominator of these companies is their definition as cybersecurity
firms. "The law doesn't allow companies or individuals to get involved
with offensive cyber," according to Dr. Harel Menashri, head of the cyber
department at the Holon Institute of Technology, who was a co-founder of the
Shin Bet Cyber Warfare Unit. "The Israeli cyber industry has made itself a
good name regarding advanced capabilities ... One of the greatest advantages
of the Israeli culture is the ability to develop and move around things very
quickly. Even if I didn't serve in the same unit with someone who I'm
interested in, I'll probably know someone who did," Menashri added. "Israelis
gain their technological knowledge during their military service through units
like 8200 and the cyber units of Shin Bet and the Mossad. That knowledge is a
weapon, and today, quite a few IDF veterans from intelligence units move
abroad and share their knowledge with foreign parties." Menshari gave the
example of a group of young Israelis who had graduated the IDF's elite Unit
8200 and a few months ago decided to go and work for the UAE-based
intelligence firm Dark Matter after being tempted by large sums of money.
How to Build an Accessibility-First Design Culture
A great place to begin is your component library. Identify which components
are used the most often and which underlying components underpin other
functions. For example, make sure buttons, inputs and links have accessible
focus and hover states. It’s a lucrative, efficient way of scaling
accessibility fixes because once you make one fix, you’ll see it propagate
throughout the organization wherever that component is used. There are a few
key factors to be aware of at this stage. First, create a clear plan for who
can make changes and how you’re testing components to ensure accessibility
features are not unintentionally removed. Second, your work doesn’t end after
creating accessible components. In the UI, individual components are put
together like puzzle pieces, and just because each piece is accessible doesn’t
mean the entire UI will be. Since the UI involves multiple components talking
to each other, you’ll need to ensure that the experience is usable and
accessible as a whole. The goal is to ensure every existing and new component
in a library is accessible by default. This way, when developers pull features
into their work, they’ll know with certainty it’s designed to be accessible.
Get it right once, and you get it right everywhere.
Powering the Era of Smart Cities
A priority for cities in the years to come will be reducing air pollution
levels. This is already a major concern – nine in ten people breathe polluted
air resulting in seven million deaths every year, according to the World
Health Organisation. As city populations and traffic volumes boom, the role of
smart technology in tackling pollution will be crucial. While data on
emissions and congestion has been available for some time, only recently have
we been able to build a full picture of its reach and harm. Fusing data from
various sources can reveal new insights to be used to manage energy use and
minimise pollution. For example, IoT sensor technology can intelligently
detect when there is little or even no pedestrian or road traffic, dimming
streetlights autonomously and saving energy. By crunching vehicle rates in
real time, as well as pressure, temperature and humidity, air quality levels
can be accurately predicted and mapped. This provides the insight to
proactively adapt traffic controls and mitigate harm. As always, the smart
move is to analyse and adopt best practices from other cities and nations.
Singapore, for example, is generally considered to be the global smart city
leader, much due to significant government investments in digital innovation
and connected technologies.
The future of tech in healthcare: wearables?
IoT and wearable devices are ideally placed to transform the management of
both preventable and chronic diseases and represent a big opportunity for
digital to disrupt the industry. Data on human health can now be collated to a
level and scale that was never before possible, while innovations in machine
learning and adaptive algorithms provide credible predictors for the risk of
diseases. Such data gives us actionable insight, empowering us to make small
but significant changes to lifestyle habits so we may work towards living a
longer, healthier life. The opportunity, however, does not come without
challenges, and two of the biggest obstacles that must be negotiated lie in
the budgetary and the clinical. On the financial side, the system either lives
or dies depending on whether doctors have the additional time and
expertise to interpret and implement a treatment plan based on the assessment
of vast reams of data. On the clinical side, non-medically graded
user-generated data makes it challenging for a doctor to include this within
the overall treatment decision-making process. The strength of AI and machine
learning, of course, is that it can cope with large amounts of data and find
statistical correlations where they exist.
Microsoft unveils Open Service Mesh, vows donation to CNCF
Open Service Mesh builds on SMI, which is expressly not a service mesh
implementation, but rather a set of standard API specifications designed
within CNCF. If followed, the specs allow service mesh interoperability across
multiple types of networks, including other service meshes, and public,
private and hybrid clouds. The service mesh layer will be a key component of
broadly accessible, real-world multi-cloud container portability as mainstream
enterprise cloud-native applications advance, Pullen said. “Service mesh
should help that, theoretically, especially if there’s standardization of it,
but it’s going to require an interesting rework to make any Docker container
compatible with any container cluster,” he said. “It’s more than putting
something in Docker, it’s about that ability to route services in a somewhat
decoupled way.” Simplicity and ease of use was also a point of emphasis in
Microsoft’s OSM rollout, which analysts said seemed to target another common
complaint about operational complexity among early adopters of Istio. OSM, by
contrast, will build in some services that have been complex for service mesh
early adopters to set up themselves, such as mutual TLS authentication.
Understanding What Good Agile Looks Like
Agile management began as a work of passion. It was born of a fierce desire
felt by disgruntled software developers to set things right. Their Agile
Manifesto (2001) not only succeeded in its modest goal of "uncovering better
ways of developing software.” It had the unintended consequence of generating
a candidate as the paradigm for 2020 management generally. Thus, Agile
management began with exploring more nimble processes for one team, then
several teams, then many teams and then the whole organization. It set in
train the emergence of firms like Amazon and Google that not only showered
benefits on their customers and users but also, for better or worse, developed
the capacity to dominate the entire planet. As society now struggles to decide
what to do about these new behemoths, it is useful to keep their possible
flaws conceptually separate from the principles, processes and practices that
enabled them to grow so fast. We need to keep in mind what good Agile looks
like—essentially a better way for human beings to create more value other
human beings. In any established organization, a small set of fairly stable
principles (also known as mindset or management model) tends to guide
decision-making throughout the organization.
Quote for the day:
