AI can see things we can’t – but does that include the future?
“What we focus on is augmented intelligence for humans to take action [on],”
says Radtke when I raise this concern. “We are not prescribing the action to be
taken based on the insights that we get – we're trying to make sure that the
human has all the necessary intelligence to drive the behavior that they need to
drive. We're reporting facts back – this actually happened here, this is what
has happened in the past – and you can take action based on that. It's all about
driving improved safety for everyone in that area.” When I press him on the
possible human rights concern and the inevitable pushback that will arise if AI
is routinely used to pre-emptively police areas deemed as problematic, he
answers: “I think that with every technology that's ever been out there in
history there is always a way to use it for non-good. I think you have to focus
on the good that it can provide and make sure that you police the non-good
behavior that could happen from it.” This will entail some sort of oversight.
“There are consortiums out there to help drive the ethical adoption of AI
throughout the industry – we definitely keep aware of those.
RPA vs. BPA: Which approach to automation should you use?
Where BPA and RPA overlap, according to Mullakara, is the goal of eliminating
human intervention in order to process multiple automation. “The whole idea of
BPA was to remove people from the process and that's kind of what RPA is also
aiming for. In the sense of the simple workflow automation, both can do it. RPA
does it through a UI integration whereas BPA does it mostly with APIs. And you
know, automating the workflow with the systems by invoking the systems,” he
tells us. However, Taulli explains that automation really won’t get rid of
people at this point and it will be the usual suspects that will, such as
recessions. Mullakara agrees that this messaging for BPA and RPA is a common
misconception and has earned both technologies quite a bad rap. “So, what you
actually automate with RPA for example is tasks – it's not jobs. It's not an
entire job even if it's a process. It’s not jobs, so we still need people,” he
says.
All the Things a Service Mesh Can Do
Many organizations have different teams and services dispersed across
different networks and regions of a given cloud. Many also have services
deployed across multiple cloud environments. Securely connecting these
services across different cloud networks is a highly desirable function that
typically requires significant effort by network teams. In addition,
limitations that require non-overlapping Classless Inter-Domain Routing (CIDR)
ranges between subnets can prevent network connectivity between virtual
private clouds (VPCs) and virtual networks (VNETs). Service mesh products can
securely connect services running on different cloud networks without
requiring the same level of effort. HashiCorp Consul, for example, supports a
multidata center topology that uses mesh gateways to establish secure
connections between multiple Consul deployments running in different networks
across clouds. Team A can deploy a Consul cluster on EKS. Team B can deploy a
separate Consul cluster on AKS. Team C can deploy a Consul cluster on virtual
machines in a private on-premises data center.
Snowballing Ransomware Variants Highlight Growing Threat to VMware ESXi Environments
The proliferation of ransomware targeting ESXi systems poses a major threat to
organizations using the technology, security experts have noted. An attacker
that gains access to an EXSi host system can infect all virtual machines
running on it and the host itself. If the host is part of a larger cluster
with shared storage volumes, an attacker can infect all VMs in the cluster as
well, causing widespread damage. "If a VMware guest server is encrypted at the
operating system level, recovery from VMware backups or snapshots can be
fairly easy," McGuffin says. '[But] if the VMware server itself is used to
encrypt the guests, those backups and snapshots are likely encrypted as well."
Recovering from such an attack would require first recovering the
infrastructure and then the virtual machines. "Organizations should consider
truly offline storage for backups where they will be unavailable for attackers
to encrypt," McGuffin adds. Vulnerabilities are another factor that is likely
fueling attacker interest in ESXi. VMware has disclosed multiple
vulnerabilities in recent months.
5 typical beginner mistakes in Machine Learning
Tree-based models don’t need data normalization as feature raw values are not
used as multipliers and outliers don’t impact them. Neural Networks might not
need the explicit normalization as well — for example, if the network already
contains the layer handling normalization inside (e.g. BatchNormalization of
Keras library). And in some cases, even Linear Regression might not need data
normalization. This is when all the features are already in similar value
ranges and have the same meaning. For example, if the model is applied for the
time-series data and all the features are the historical values of the same
parameter. In practice, applying unneeded data normalization won’t necessarily
hurt the model. Mostly, the results in these cases will be very similar to
skipped normalization. However, having additional unnecessary data
transformation will complicate the solution and will increase the risk of
introducing some bugs.
Git for Network Engineers Series – The Basics
Version control systems, primarily Git, are becoming more and more prevalent
outside of the realm of software development. The increase in DevOps,
network automation, and infrastructure as code practices over the last
decade has made it even more important to not only be familiar with Git, but
proficient with it. As teams move into the realm of infrastructure as code,
understanding and using Git is a key skill. ... Unlike other Version Control
Systems, Git uses a snapshot method to track changes instead of a
delta-based method. Every time you commit in Git, it basically takes a
snapshot of those files that have been changed while simply linking
unchanged files to a previous snapshot, efficiently storing the history of
the files. Think of it as a series of snapshots where only the changed files
are referenced in the snapshot, and unchanged files are referenced in
previous snapshots. Git operations are local, for the most part, meaning it
does not need to interact with a remote or central repository.
Deep learning delivers proactive cyber defense
The timing couldn’t be better. The increasing availability of
ransomware-as-a-service offerings, such as ransomware kits and target lists,
are making it easier than ever for bad actors—even those with limited
experience—to launch a ransomware attack, causing crippling damage in the
very first moments of infection. Other sophisticated attackers use targeted
strikes, in which the ransomware is placed inside the network to trigger on
command. Another cause for concern is the increasing disappearance of an IT
environment’s perimeter as cloud compute storage and resources move to the
edge. Today’s organizations must secure endpoints or entry points of
end-user devices, such as desktops, laptops, and mobile devices, from being
exploited by malicious hackers—a challenging feat, according to Michael
Suby, research vice president, security and trust, at IDC. “Attacks continue
to evolve, as do the endpoints themselves and the end users who utilize
their devices,” he says. “These dynamic circumstances create a trifecta for
bad actors to enter and establish a presence on any endpoint and use that
endpoint to stage an attack sequence.”
Towards Geometric Deep Learning III: First Geometric Architectures
The neocognitron consisted of interleaved S- and C-layers of neurons (a
naming convention reflecting its inspiration in the biological visual
cortex); the neurons in each layer were arranged in 2D arrays following the
structure of the input image (‘retinotopic’), with multiple ‘cell-planes’
(feature maps in modern terminology) per layer. The S-layers were designed
to be translationally symmetric: they aggregated inputs from a local
receptive field using shared learnable weights, resulting in cells in a
single cell-plane have receptive fields of the same function, but at
different positions. The rationale was to pick up patterns that could appear
anywhere in the input. The C-layers were fixed and performed local pooling
(a weighted average), affording insensitivity to the specific location of
the pattern: a C-neuron would be activated if any of the neurons in its
input are activated. Since the main application of the neocognitron was
character recognition, translation invariance was crucial.
Don’t Just Climb the Ladder. Explore the Jungle Gym
Most of us do not approach work (or life) with a master plan in mind, and
many of the steps we take are beautiful accidents that help us become who we
are. “I’m 67 years old,” Guy said, “and I think I finally found my true
calling.” He was referring to his podcast, Remarkable People, where he
interviews exceptional leaders and innovators (think Jane Goodall, Neil
deGrasse Tyson, Steve Wozniak, and Kristi Yamaguchi) about how they got to
be remarkable. “In a sense, my whole career has prepared me for this moment.
I’ve had decades of experience in startups and large companies. So that
gives me the data to ask great questions that my listeners really want the
answers to,” Guy said. Guy is undeniably brilliant, and his success is no
accident. But still, he believes that luck has played a part in his success.
In his words, “Basically, I’ve come to the conclusion that it’s better to be
lucky than smart.” Maybe Guy is right. Or perhaps, the smartest people know
when to take advantage of luck and act on the opportunities that present
themselves. Whatever the case, it’s important to take calculated risks.
Should You Invest in a Digital Transformation Office?
With the digital transformation office comes a transformation team, who
initiates organizational change. Laute says that it’s crucial that everyone
inside the organization stand behind the transformation team if they truly
want to see changes happening. “You need to have an environment where these
people, the transformation lead and the transformation team, are allowed and
are not afraid to speak up. These people shouldn't be biased, not just
following what the executive board says, but really [being] able to
challenge and to speak up. And they should have the freedom to call out if
something is going in the wrong direction, may it be content or
behavioral-wise,” she explains. And while clearly there can be
technology-related challenges, Laute tells us that digital transformation is
also a people problem, and calls for a change in culture and mindset in
order to find success. The cultural shift, she explains, is truly where
everything starts to come together in order to get the transformation going.
“Digital [transformation] is not only technology. You need to change
behaviors and you need to change processes. And most of the time, you change
your target operating model, right?”
Quote for the day:
"Uncertainty is a permanent part of
the leadership landscape. It never goes away." --
Andy Stanley
No comments:
Post a Comment