So, hi everybody.
Uh, today I'm gonna tell you about Guardian Agent.
A tool we've built for Secure Delegation for the SSH protocol.
And towards the end of the talk,
I'm also gonna [NOISE] discuss how this
tool we've built as well as the principles behind it,
uh, allow us to address other authorization scenarios,
and and not only SSH.
[NOISE] But we're gonna start with SSH,
and we're gonna see how does- uh,
to see how delegation works today in SSH.
Let us consider the following scenario
that many of you are probably gonna be familiar with.
So, imagine a user connecting to some,
er, Cloud Machine of her choice,
and that user wants to clone her private Gi- to- Git repository located on GitHub.
And the repository being private of course,
[NOISE] uh, cloning this repository,
requires the user to authenticate.
[NOISE]. Uh, but the private key of the user is
located on her local machine on her laptop.
And obviously she would like to avoid just u- uploading, and,
an- uploading her key to those,
to this Cloud Machine which could be potentially untrusted.
While this seems kind of,
um, like a kind of basic scenario,
and i- it also seems like there're SSH protocol,
and the SSH tool chain are kind of a,
kind of thitherward, and we have a solution for that.
So, what can the user do?
[NOISE] Um, the user can run, [NOISE] uh,
can run an SSH agent locally,
and add her key to that SSH agent.
And to be extra secure, she's also gonna, uh,
add this minus c flag that is going to ask that
such that the agent is going to ask her every time her key is getting used.
So she'll know exactly, uh, how the key is used.
Then, [NOISE] uh, the user is gonna, uh, SSH,
and enable agent forwarding,
uh, when she connects to the Cloud Machine.
And then, when she runs that,
uh, when she runs the git clone command, uh,
the authentication request is going to get forwarded, uh,
to the local agent, who's going to prompt the user to confirm the use of her key.
And it looks like problem solved.
So the user knows every time her key is used,
she can confirm it, she can deny,
she has full control of her credentials.
But, if you look closely at this prompt to the user, er,
that the users get you kind of see that it doesn't
really provide us with the guarantee that we would want.
So what does the user see on this prompt basically,
she sees the path to a local key not really the kind of
information that you want to have once you're confirming the operation.
And so, when the user clicks "OK",
and she think- she thinks she authorizes that the Cloud Machine to clone her repository,
if that Cloud Machine is compromised or maybe the entire Cloud provider is compromised,
then, well she thinks she's signing, uh,
a challenge use for authentication to GitHub that malicious Cloud,
Cloud provider could actually use that sign authentication challenge in order to
perform some different malicious operation on
a completely different host to which the user has access.
And what is kind of the root cause for- for this problem?
Well, the root cause is that with the current way that this protocol, uh,
is used and all the tools around it are built that the user can clearly verify
the identity of the delicate- the delegate to whom she delegates her credentials.
So, when she see- sees this prompt she doesn't
really know that this challenge that originates from one,
from, from GitHub, or does it originate from some, some different host.
And also the user cannot really verify the command that, uh,
that, uh, this authorization is going to be- is going to be used for.
And the identity of the server and the delegate, both,
are kind of untied, uh,
when you- when she sees that, uh, authentication challenge.
And it's important to stress that this is not merely a UI issue.
So, it's not just we could change that prompt and display some other uh,
maybe more important information.
In fact this type of information isn't really available to the client.
And of course there are
no any cryptographic guarantees that when you when- you signed this challenge,
what does this challenge going to be used for?
So, the picture today is that by approving this,
uh, approving these kind of prompts,
we are essentially signing a blank check.
So, we put our signature there, uh,
and we don't even specify the amount of money,
what's the command that we're going to run.
We don't specify the, the recipient,
the identity of the server on which the command is going to run,
and maybe even the identity of like the bank if I have multiple bank accounts is
also not set to relate to the previous stock.
So, what do we really want?
So what we want is instead of- [NOISE], uh,
instead of having this kind of not very useful prompt that doesn't allow us to
control the delegation what we wanna have is- we wanna have uh,
uh, we wanna have an authorization decision that has all the scope,
all the necessary scope to make a meaningful decision.
So at a minimum,
we would like to know who are we delegating to?
What can the delegate do with our identity and to whom.
So for example, what's the in- what's the server to which the operation will be applied.
And again I want to stress out that,
it's not kind of a- it's not on the UI issue but kind of uh,
protocol [NOISE] there is a protocol issue here as well.
Um, so, and if you think that this is kind of an issue that's only,
only present on SSH and other protocols get it right.
Well, similar scenarios is not exactly of this type,
but similar scenarios involving, uh,
unscoped delegation exist in other protocols as well.
So for example, for PLS,
we can consider this scenario in which some publisher wants to
delegate the duty of serving its home page to a Cloud,
uh, to a Content Delivery Network, a CDN,
and of course the publisher wants to still serve his,
uh, serve his homepage, uh, using HTTPs.
So, to do that, uh,
currently, what is required is that for the publisher,
it needs to give the certificate as well as
the private key corresponding this certificate to the CDN,
uh, for the user to get an HTTPS connection.
But once the publisher
gives access to its private key to the content delivery network,
if the content delivery network is compromised, or is malicious,
then he can use, um,
he can use this credential to serve some other content.
So, in a- in- in essence,
the publisher delegates his identity without,
uh, without sufficient control.
And there have been like rec- there have been recent works, for example,
the work by CloudFlare called KeylessTLS, which is er,
which in essence is very similar to SSH agent and suffer from similar problems.
So, with KeylessTLS,
the solution works by the publisher keeping the private key.
So in this picture you can see that the private key is kept,
is kept by the publisher.
And then, every time the, the CDNs,
every time that the CDN wants to establish
a secure HTTPS connection as the server, he's gonna,
uh, uh, er, he's going to talk to the publisher,
and send him th- send him a challenge,
and gonna get a signature or decryption depending on the,
uh, uh, encryption protocol used.
We- so this- so if you look at this solution in essence,
the reason that I said it's kind of similar to SSH agent is that
the publisher really exposes a signing oracle here.
So, he tells the CDN just send me those challenges and I'm gonna sign them.
Do I know how they're,
gonna use it. Not really.
So if the CDN really wants to manipulate the content, uh, he can do that,
and for example, the identity of the client is also not verified,
um, in this case.
So, in our work, we started off by addressing the SSH issue and rebuild Guardian Agent.
Guardian Agent is a tool, uh,
that enables secure delegation for the SSH protocol.
Our tool consists of an authentication agent running on the user's,
uh, local machine on her laptop.
Um, it consists of a custom SSH client,
uh, that you can run on the- your- on your cloud machines,
or any other partially trusted machine on which you wanna run,
uh, operations using SSH but you don't want to put your key on.
And the nice thing is that our- our protocol and
our tools are backwards compa- compatible with existing SSH servers,
so you don't really need to modify the SSH server.
So, you don't really need to go and- and wait for GitHub to adopt our solution,
and you can get practical,
uh, security benefits today.
Uh, so let me tell you about how does Guardian Agent work.
Um, so in the setting we have
the local machine that keeps the private key and running our agent,
we have the delegate running the custom client,
and we have GitHub as the server.
So, er, the first thing- the fir- uh,
when the delegate wants to run the command,
uh, he's gonna- instead of running the command directly on the server,
he's gonna talk, uh,
to the local machine through- through
the Guardian Agent protocol and request the agent, uh,
to run the command, uh,
and specify which server,
uh, he wants to run the command on.
Then the agent can prompt the user and make sure that the user actually,
uh, that the user actually approves the operation.
And in this- in- in this way the, uh,
the Guardian Agent, uh,
local machine gets the- the command,
gets the identity of the server and can also identify,
um, ide- identify the delegate based on the incoming connection.
Then if the user approves the operation, uh,
the local machine starts a connection to the server,
authenticates using the private key, and issues the command.
So, uh, at least we can see that
the local machine has control over
the- over all- all the parameters that we want to control,
the identity of the server and the,
uh, and the command and so forth.
But this seems kind of,
um- kind of, uh,
naive in- in the sense that if we just do it
this way we're kind of losing all the benefits,
uh, for which we wanted to delegate in the first place.
So, we really want it to run the command on the delegate,
and we don't want to run on the local machine for whatever we wanted to do,
and even tunneling the entire connection
through the local machine isn't really practical,
because imagine you were just running on-
riding on a train and you have a ve- very bad connection,
you don't really want to tunnel
the entire Git repository you're cloning through your local connection,
you just want to get it from the server to the delegate.
So to do that- what,
uh- the final step of our protocol is basically handing off the SSH connection
from the local machine to the Git- to
the delegate in a way that it's transparent to the server.
This way only the first part of establishing the connection which is really, uh,
minimal in terms of the required bandwidth
happens between the local machine and the server,
and then the connection can be resumed directly between
the delegate and the SSH server so the entire contents of this,
let's say, Git- Git- repository, in this case,
will be transferred directly from the server to the delegate.
Um, so, uh, how do we do this,
uh- how do we do it- do it this SSH handoff?
And to kind of, um,
think of it abstractly what we really want to do is we want to
serialize the entire state of the SSH client,
send it to the delegate and then de-serialize and resume the connection.
But if we look at what does the SSH state consist of,
it contains a- a lot of messy details.
It contains, uh, of the session ID and
the sequence number and all the details of the, um,
uh, of- of what command I'm running and most annoyingly,
it has a lot of crypto details.
So, every time you establish an SSH connection you need to negotiate an algorithm for,
um- for key exchange,
and for MACing, for encryption,
and every- each such algorithm contains a lot of
state that evolves as the connection continues.
You, uh- you have IVs, you have counters,
and just serializing this in general would require either support from
each individual cipher or has to- have- or having to fix the cipher mode,
and we don't really want to do that.
So, to avoid doing this kind of messy serialization what we do is we use, um,
a feature already provided to us by the SSH protocol and
this is that we can we- we can run a key re-exchange on an existing connection.
And the nice thing about key re-exchange is
as you can see here is that all the crypto state is
actually refreshed when you'd- when you- when- when
you do the key re-exchange which is sort of what we need.
Um, so how do we do- how we- do we use this, uh, key re-exchange?
So, we start by having two separate SSH connections,
one between the delegate and the local machine,
and the second one between the local machine and the server.
And initially, those two connections are independent.
Then, uh, after the local machine has sent the command, uh,
o- on its connection with the server, uh,
and a key re-exchange is triggered on both connection.
But this time, the key- the parties in
the key re-exchange are the delegate and the server,
and the local machine only acts as a relay that passes around,
uh, uh, the key exchange messages.
And kind of as a nice,
a- a- as a basic property of key re-exchange,
what you end up having is you end up having a shared,
um- a shared key between the two endpoints namely,
the delegate and the server.
So, once we're done, uh, once we're done with the key re-exchange,
we essentially have, uh,
an SSH connection now between the delegate and the server.
And there are a couple of mo- a couple of-
the messy details that you need s- still need to take care of,
uh, but they are very minimal and don't depend,
for example, on the crypto being used.
And one last thing that needs to be taken care of is that we want to avoid the delegate,
uh, being able to issue additional commands after the handoff.
So, to do that, before we initiate the handoff we use, um,
an already existing, uh,
extension to the SSH protocol,
uh, available for, uh,
in OpenSSH called no more sessions which allows the local machine to, uh,
um, to request from the server not
to honor any additional commands that's gonna be, uh, issued after that.
This essentially locks down the session to the particular command.
And, uh, maybe- maybe something that, uh,
jumped to your mind when- while I was describing this,
where we- we also have a couple of constraints on
the transport layer that we need to take care of.
For example, the agent might not even have direct connectivity with the server,
because maybe you're connecting to the delegate and it needs to connect to some,
uh, server which is not accessible from your local machine.
And second, as I've told you we want this to
become fully compatible with existing servers,
um, uh, so the entire connection- thi-
this entire handoff needs to be transparent from the server's point of view.
So, to address both of these concerns,
what we do is that we- the delegate acts as
a relay on the transport layer between the local machine and the server,
and then after the handoff he can reuse
the same TCP connection to talk to the server directly.
So, jis- this achieves both benefits.
So, this really allows the local machine to have connectivity with the server,
and also from the server's point of view, uh,
he- throughout the entire connection,
he talks over the same connection and it looks transparent.
Um, so Guardian Agent, uh,
uh- we implemented Guardian Agent,
it is available and already used,
um, by people online.
So, you can- I- I invite you to go on GitHub and use it.
[NOISE]
Um, and kind of the- the main part of the talk I want
to kind of discuss about how Guardian Agent both as a tool and those fine-grained,
uh, fine-grained delegation principles can actually be extended to additional settings.
And specifically, I'm gonna talk about, uh,
pri- uh, I'm going to talk about talking pri- privilege elevation.
So what's privilege elevation?
So I think most of you are again familiar with the need to
run the sudo command for example to install some program.
And the reason that you need to do it is,
well, you usually run, uh,
using some un- un- unprivileged account and you have some operation that you
want to do which requires more privileges.
And so again sudo is kind of ubiquitous.
You run it and then it asks you for your password and once you confirm your password,
uh, the entire process runs as root.
So what's the ki- what's the problem with this?
So there are kind of two issues with it.
The first issue is that, uh,
if- if you- if you hope that sudo would allow you to have
some security separation between your unprivileged account and your privileged account,
then it doesn't really, uh, it doesn't really do that.
And specifically, it doesn't really withstand an atta- an attacker
that manages to compromise the unpriv- the unprivileged account.
Such an attacker can, for example,
wait for you to run sudo for- for
some legitimate cause and then try and phish your password,
modify the command and this can be done using like
very basic modifications even to your shell script.
Um, so sudo is still good for preventing yourself from making a mistake,
but it's not- I- I- I kind of claim that it's not really,
um, suitable to give this extra security,
uh, extra security guarantees between your unprivileged account and the root account.
And second, if you look at the way that this mechanism work,
it's that after you approve, uh,
after you approve the privilege elevation and even if ki- the program,
uh, to which you gave these elevated privileges is ah- a legitimate program,
it is in fact the one that you want to- the one that you want to run,
you still kind of give it a- you- you give it
full root privileges that are unrestricted in scope.
And so you're allowing it to do anything that root can do rather than the-
perhaps a single operation that you wanted to give it these extra pri- privileges for.
And this is dangerous mostly because programs can have, ah,
programs can be exploited either due to memory bugs or some logical security bugs,
and whenever this privilege- you basically increase your attack surface
because whenever this priv- whenever this privilege process gets compromised you lose f-,
uh, you lose full control of your machine.
So, um, our idea is kind of to
use the same principles behind Guardian Agent to solve this,
uh, to solve this problem or maybe not to solve,
but kind of to improve the state of the world,
uh, when it comes to privilege elevation.
Um, so how do- how- how are we solving this?
So, again consider the scenario where we have a local machine, and the one that, uh,
has your credentials, the one that you use locally,
and now consider a case when you want to run sudo on some remote machine.
And I- I wanna know that the same kind of principles can be applied
to running sudo even on your local machine but then you need to take care of, uh,
trusted path issues, uh,
such as you need still to- you still need to have somewhere where
you can enter your password or interact with the machine,
uh, in a secure way.
But for simpli- simplicity,
let's just look at the remote scenario.
So, uh, with our- with our solution,
instead of running the apt-get binary using sudo and elevating the entire process,
the process is going to run, uh,
using the normal user privileges.
So this obviously won't work.
So what do we do, um,
so what we do is that we run a separate privileged daemon that
will take care of elevating privileges in a mine- much more fine-grained way.
And we also have a small library running within,
uh, the unprivileged- unprivileged process.
So everything on the left here, uh,
the apt-get process and the library is unprivileged and we have one privileged daemon.
So when the apt-get process is going to, uh,
do some operation which requires those extra privileges, the one,
uh, that we actually originally needed sudo for,
then obviously this operation is going to fail.
So, for example, uh, in this case,
it's going to try and open some fi- protected file which you cannot write as
an unprivileged user and operating system were
gonna tell- is going to tell the process, well, you can't do that.
Access is denied. Uh, in this point,
our library is going to intercept that system call just before it returns to- to
the program and request
the authentication agent running on your machine
to approve this operation which was originally denied.
Uh, again, same as before now the user can be prompted.
He can make sure that he's approving whatever he, uh,
whatever he wants, uh, to be performed on that remote machine.
And once the operation is approved,
a signed approval using the user's private key is sent,
uh, from the local machine, er,
to this library running in the unprivileged process.
Now, this unprivileged process can present the signed approval to
the privileged daemon and request for this operation to be performed on its behalf.
The privilege daemon verifies the- verifies the signature,
verifies the parameters of the operation,
and then goes- goes ahead and performs this operation from the privileged context.
Er, in this case, for example,
it opens the file,
gets the file descriptor and passes the file descriptor back to the unprivileged daemon.
So as you can see here, uh, as a result,
we've managed to perform a privileged operation from an unprivileged process without, uh,
first of all, by allowing the user to control this specific operation and
without increasing the attack surface and running the entire process as root.
Um, so this is kind of what we're working on now.
And it turns out that kind of
this personal auth- authentication agent that we're building opens also,
uh, new opportunities for other things to be done.
So once we have this one focal point for all authorization decisions,
uh, for the user,
this allows us to, uh,
look for other new, uh,
possibilities for example, uh,
running a personal authorization log that will have, um,
that will have a log of all operations that- that were
performed users- using the user's identity and allow,
uh, allow the user to ask questions such as,
how was my identity used?
What privilege operations did I approve?
And make sure that his identity wasn't misused.
And second, once all, uh,
once all operations on behalf of the user kind of are
strongly tied to the user's cryptographic identity,
to his private keys located in one place,
we can then go and, uh,
maybe thi- think about other scenarios such as splitting the key between the user's, uh,
hardware token and some trusted server, uh,
in a way- in a way to enable, uh,
things like- such as two factor authorization and so on.
So, kind of to conclude, uh,
I thi- I think that the main point, um,
that you should take away from this talk is that, um,
we want to give user the power to really control the way that their identity is used.
Uh, not only- like to control every oper-
every operation individually with a much finer scope.
And Guardian Agent is a step towards this.
We provide secure, uh,
delegation for SSH and also secure privilege elevation.
Thank you.
[APPLAUSE]
OAuth 2 is becoming quite popular and it's
an authorization framework rather than an authentication one.
Um, how would that affect the- the kind of thing you're doing?
Um, so I think for OAuth 2 it doesn't really use,
um, it doesn't really use cryptography to assert every,
er, every authori- every authorization operation that you do.
So you kind of need to, uh,
again trust some- some third party to- to store your, uh,
store your tokens and- and so on.
Okay.
Any other questions?
How do you handle for- for privilege escalation?
You have to deal with states on the server.
So, for example, if you have- if you have a symbolic link
and the user approves access to a symbolic link,
but then the link changes,
how do you- how do you- how do you handle things like that?
So, actually, symbolic link is- is really a source of,
uh, kind of trouble, but, uh,
what we do is we try to dereference
all- as much of the state as possible from the unprivileged context.
Uh, so for example, all symbolic links or relative parser
are evaluated in unprivileged process, and then,
uh, when the user approves this- approves the path he
can get- he gets an absolute path that he can assign some trust to.
Cool, yeah.
Any more questions?
Isn't there a kind of inherently, uh,
time to check time to use problem there even with dereferencing all these symbolic links?
Um, yes.
I think one idea that we have to address,
this kind of issues is for example, uh,
maybe not blocking, but alerting the user if any component
of the path let's say for file operation is writable by the unprivileged account.
Uh, and you can kind of think of an environment if all the components of
the path are not writable by the user, uh,
then it's not to- it shouldn't be susceptible to this kind of attack at least
from- There's- there's still
global state whenever- whenever you do which can be modified,
uh, concurrently, um, but I think it can minimize the risk of that.
How about situation how user is
allowed to perform actions on Machine A and not on Machine B.
So, if- if- if user is allowed to say run apt-get on Machine 1 whereas she's,
uh, I mean she's but then
she's not allowed to run on Machine B because it's only command?
Um, so from each machine keeps, um,
keeps a set of public keys of- that
corresponds to the users that are allowed to do this operation.
This is from the server's point of view and
from the agent's point of view when the user authorizes something,
it's actually presented also with identity of the s-.
So in the example that I gave you,
you will not only see that you are approving
apt-get but you also see the identity of the machine.
And this identity is going to be- to be inside
that- that blob of data that you are going to sign.
And then each machine can verify that its own identity was signed by the user.
How does the- the mechanism seems a lot like what Windows does.
Um, how does the user actually
know the answer to the question that the user is being asked?
Um, so- so I think there's like- you're bringing up two- two important issue.
First of all, I think that if you compare it to let's say the UAC mechanism in Windows,
so it- it- it- although you are given some prompt,
it still suffers from the problem that
the entire process gets elevated after you approve these things.
So you don't if- you still suffer from increasing the attack surface.
About the point that the user needs to make a decision, uh,
then kind of our hope is that on the agent side using the policy we can- we can kind of,
um, uh, present the question in a meaningful way.
For example, uh, not only show the full path,
but give some meta information about it.
I guess the answer
that I sort of sourced to my question was actually the authorization log.
That you could at least forensically go back and
say although you didn't understand that you were authorizing this,
you did actually authorize this and this was
the thing that caused you to be broken later, would be improved.
That's fine.
Yeah.
Great. Thanks Dima. [APPLAUSE]
For more infomation >> A Guardian Agent for Secure Delegation - Duration: 26:45.-------------------------------------------
Hearing for mom accused of drowning daughter - Duration: 1:18.
For more infomation >> Hearing for mom accused of drowning daughter - Duration: 1:18. -------------------------------------------
7 Amazing Benefits Of Baking Soda For Hair Skin And Body | Benefits Of Baking Soda On Skin - Duration: 3:01.
baking soda acha sodium bicarbonate is a key
ingredient used in baking fluffy and delicious cakes this baking staple helps
to neutralize and regulate the pH level of a substance and ensures that it is
neither too alkaline nor too acidic today's video will discuss why you
should add baking soda in your daily life before you watch this video please
take a moment to subscribe our youtube channel by clicking the subscribe button
then tap the bell icon so you will be the first to know when we post new
videos daily one use it as a toothpaste mix a teaspoon of baking soda with 1/4
teaspoon of hydrogen peroxide and then brush your teeth it can be used as an
alternative to commercial toothpastes for an extra boost you can dip your
toothbrush along with paste and baking soda and get sparkling teeth to
deodorize your lunchbox freshen up the sour smelling lunchbox and bags using
baking soda open the lunchbox and sprinkle a heavy layer of soda in it
keep the lid open overnight wash it with hot water and dish soap the next day 3
use it as a carpet freshener baking soda is one of the cheapest natural carpet
freshener sprinkle this baking staple on the carpet leave it for a few hours and
vacuum up your carpets will now smell awesome 4 freshen up the closet place a
box of baking soda on the Shelf of your closet and keep your closet smelling
fresh 5 use it to wash your car now clean the lights of your car windows
tires seats and floor mats using baking soda mix 1/2 cup of baking soda in half
bucket of warm water and use the solution with a sponge or soft cloth to
clean your car 6 as a homemade deodorant prepare of deodorant by mixing equal
parts of baking soda and cornstarch combine the ingredients with tea tree
oil and store it in an airtight container
apply the mixture onto your underarms and neutralize the body odor 7 has a
fire extinguisher baking soda can help in handling minor kitchen fires
electrical or grease just throw a handful of baking soda at the base of
the flame baking soda when heated produces carbon dioxide that helps to
put out the fire don't forget to call the fire department just to be on the
safer side have you ever tried baking soda for you daily use let me know in
our comment section below if you liked this video give it a thumbs up and share
with your friends for more daily tips subscribe to our channel below
thank you
-------------------------------------------
Welcome to SupportAssist for PCs - Duration: 1:13.
Thank you for choosing Dell!
I'd like to introduce you to SupportAssist, created by Dell,
to be your single source for PC help.
SupportAssist is like having your very own Dell expert
inside your system, here to help whenever you need it.
You can find it on your tool bar or by searching
"SupportAssist" in your Windows start menu.
SupportAssist allows you to check your
PC's hardware and software health
and provides update recommendations when needed.
It helps you request support, find service plan details
and keep drivers up-to-date.
Does your PC appear to be running slow?
Click "Run All" button, to automatically refine
PC settings.
All of this is included with the purchase of your Dell PC.
To help us serve you faster,
complete your profile within SupportAssist.
And if you ever need support, all you have to do
is confirm your information.
For even more support, unlock additional features,
such as automated virus removal,
by upgrading to a Premium Support Plus warranty.
If you have a problem or need assistance,
Support Assist is always here for you.
-------------------------------------------
DeSantis Draws Criticism For Remarks - Duration: 2:13.
For more infomation >> DeSantis Draws Criticism For Remarks - Duration: 2:13. -------------------------------------------
CONGRESS DROPS SURPRISE NUKE ON OBAMA – HE'S CUT OFF FOR LIFE! - Duration: 11:40.
CONGRESS DROPS SURPRISE NUKE ON OBAMA – HE'S CUT OFF FOR LIFE!
Barack Hussein Obama has been living the high life since he left office, using his six-figure
taxpayer-funded pension to take as many vacations as he possibly can.
Now, however, his free ride on the dime of American taxpayers may be over for good.
Western Journal reported that the Senate Committee on Homeland Security and Governmental Affairs
has approved a bill that would save taxpayers millions of dollars by cutting presidential
pensions.
Introduced by Republican Iowa Sen. Joni Ernst, the bill would cap the pensions at $200,000,
with adjustments made for the increase in the cost of living each year.
The bill also makes it so that for every dollar above $400,000 that a president makes from
things like speeches and other post-presidential events, one dollar would be taken off their
pension.
Obama has been making over $400,000 for numerous individual speeches since leaving office.
"Our national debt now exceeds $20 trillion; this bipartisan effort is another important
step toward reining in Washington's out-of-control spending," Ernst explained.
Obama has just signed a hefty deal with Netflix along with his wife Michelle that is said
to be worth millions, so he is not going to be short on cash anytime soon.
"One of the simple joys of our time in public service was getting to meet so many fascinating
people from all walks of life, and to help them share their experiences with a wider
audience," the former president said in a statement.
"That's why Michelle and I are so excited to partner with Netflix – we hope to cultivate
and curate the talented, inspiring, creative voices who are able to promote greater empathy
and understanding between peoples, and help them share their stories with the entire world."
There's no reason why taxpayers should have to foot the bill for the Obama family's
lavish lifestyle when many of us are struggling just to put food on the table thanks
to the former president's ridiculous policies.
-------------------------------------------
Raw Diets for Pets - Duration: 3:22.
For more infomation >> Raw Diets for Pets - Duration: 3:22. -------------------------------------------
Police Search For Manhattan Beach Rape Suspect - Duration: 0:45.
For more infomation >> Police Search For Manhattan Beach Rape Suspect - Duration: 0:45. -------------------------------------------
Race Day For The Supercar Family - Duration: 11:54.
For more infomation >> Race Day For The Supercar Family - Duration: 11:54. -------------------------------------------
Final day of public visitation for Aretha Franklin held Thursday in Detroit - Duration: 4:33.
For more infomation >> Final day of public visitation for Aretha Franklin held Thursday in Detroit - Duration: 4:33. -------------------------------------------
McCain family at State Capitol for procession to North Phoenix Baptist Church - Duration: 8:57.
For more infomation >> McCain family at State Capitol for procession to North Phoenix Baptist Church - Duration: 8:57. -------------------------------------------
WBZ Midday Forecast For August 30 - Duration: 3:41.
For more infomation >> WBZ Midday Forecast For August 30 - Duration: 3:41. -------------------------------------------
Dharam Taylor Has Been Awe with Mercedes-Benz for Decades | #MyMBEW - Duration: 1:37.
Currently I'm in the Mercedes-Benz S-Class
this specific one is the Mercedes-Benz S560
and I just wanted to tell you
my first memory of being in a Mercedes-Benz.
It just so happened to be in a S-Class
my uncle was driving it, it was actually his friends
and I still remember the number plate on it
It was a personalized number plate
in London, England
It was my uncle's bosses name
and it was white just like this one
and the embody memory that I have is kind of quirky
other than it being luxurious and refined
beautiful!
Were in fact the headrests
and my memory of the headrests are them going up like this
and just being fascinated
I was in the back seat at the time
I couldn't have been older than 12 or 13 years old
and I was just in awe
about how the headrests looked
and he was just doing it automatically
and that's because any time that I'd seen a headrest prior
I'd have to do it manually
I thought that was quite funny!
I'm Dharam Taylor here at Mercedes-Benz Edmonton West
Comment and let me know what you think!
-------------------------------------------
Police Search For 2 Suspects Wanted In Violent Robbery In South Philadelphia - Duration: 0:22.
For more infomation >> Police Search For 2 Suspects Wanted In Violent Robbery In South Philadelphia - Duration: 0:22. -------------------------------------------
Bat found at Pasadena shopping center tests positive for rabies - Duration: 0:23.
For more infomation >> Bat found at Pasadena shopping center tests positive for rabies - Duration: 0:23. -------------------------------------------
Season wrapping up for the Charleston Riverdogs - Duration: 1:58.
For more infomation >> Season wrapping up for the Charleston Riverdogs - Duration: 1:58. -------------------------------------------
Technology Innovation for Healthcare in Aging Societies - Duration: 1:12:13.
>> From the Library of Congress in Washington, D.C.
>> Sandra Charles: Good morning and welcome to this lecture
on [inaudible] Telemedicine,
but with a whole lot more girth to it.
We are pleased to have Professor Auffret here from --
talking to us about artificial intelligence
and we're hearing so much.
I get at least three or four emails a day talking
about different aspects both of AI and Telemedicine,
just all the different terminology
and all the different ways in which this is being used
in health services and in aging services in particular.
So I was more than happy to join Dr. Steen
and the Science Division in bringing Professor Auffret here
and I know you're in for a treat.
We'll be all so much more edified when this is over
and be able to speak about the whole topic.
And not only that, but we'll be more encouraged and energized
to go do more research or read more about it
because it's not a passing fad.
It is where we are and where we are going.
So I will ask Dr. Steen to introduce our speaker
and I'd say sit back and be ready for a wonderful program.
Thank you so much.
I encourage you -- this is actually a part of our Health
and Wellness Programming for the month of May, and it's sort
of the opening lecture which we always dedicate
to some aspect of aging and health.
And then next door will be the various providers
that you can ask questions on a variety of things.
We have several area hospitals participating as well
as individual pharmacies and providers.
So we hope this day will prove for you
to be much more energizing than you've had in a while in terms
of thinking about and taking care
of your health and wellness.
So thank you so much for being here.
I failed to introduce myself at the beginning.
For purposes of the program I'm Dr. Sandra Charles,
Chief of the Health Services Office
and also Chief Medical Officer here at the Library of Congress.
Thank you again and welcome.
>> Thank you.
>> Tomoko Steen: It's a great pleasure
to [inaudible] Professor Auffret again.
I'm Tomoko Steen, Science, Technology,
and Business Division of here at the Library of Congress.
Professor Auffret last year gave a very interesting talk
on cybersecurity.
And this was recorded, so, you know, please let me know.
I'm happy to show you a URL
and you can watch the talk as well [inaudible].
And we are so lucky to have Professor Auffret twice,
and he is a Research Partnership Director at School of Business
as well as he's Associate Director for the Center
for Assurance Research and Engineering
at the Engineering School.
So he has undergraduate degree from Duke University.
He's A.B. Duke [inaudible] and also MBA
from the University of Virginia.
And he's also PhD in Physics from American University/
So he has just broad range of background.
And he has been Director for the MES Programs
at the George Mason University.
And also he's charity ranges quite broad from CIO, ICT,
cybersecurity, so the team explained what's CIO and ICT,
but it's so many, you know,
discussion right now you're going to, you know,
[inaudible] our aging population
and this is a very important topic.
So Professor Auffret is also [inaudible]
for Healthy Information Technology Advisory Committee,
and also worked with the World Bank and variety
of important organizations including World
Health Organization.
And also he was a business executive
for the MCI and British Telecom.
He has also appointment with Duke University Center
for International Development.
And he is a Physicist in Residence
at American University.
So ICT. May be I just -- so many people ask me about what ICT
when puts announcement.
It's Information and Communication Technology.
And, you know, many countries were facing such issues
of aging, how to manage aging population.
So today's lecture is going to review and analyze and status
in the potential of ICT for healthcare in aging populations.
And it's a wonderful topic to discuss and he has a broad range
of connection internationally so that his topic is going
to cover America's, Europe, Africa, and Asia.
So before [inaudible] please join me
in welcoming Professor Auffret.
>> Jean-Pierre Auffret: Thank you, Tomoko.
[ Applause ]
Where did my pen -- thank you.
Thank you all for joining and thank you Dr. Charles.
As Tomoko said, the talk is going to range from healthcare
to aging society to technology to policies.
And policies help introduce technologies
into a society for healthcare.
So it will be quite a broad range, and I'm looking forward
to the discussion afterwards, including from a special guest,
Yuki Memora [assumed spelling], who's also very active
in the aging society topic
and hosting a Washington Innovation Network here
in Washington.
So, welcome, Yuki.
Shoot. Just kind of to review then,
going to have some background on aging society
and the technology landscape, some of the challenges
in what's actually needed to spur innovation in the field,
and then a little look, short look, at National ICT Policies
and National Aging Policies and Strategies.
Before we start, I wanted to play a short video.
Just have a couple of these from Japan on some
of the introduction of technology
into assisted living facilities.
So a range of experimentation in Japan,
both with the rapidly aging society looking to further care
for seniors but also to be a world leader
in silver technologies,
to be able to export these technologies
to other countries around the world.
We'll play this short video on some of the applications
at this one nursing home,
and I think it gives a pretty good feel
for the range of possibilities.
[ Inaudible Conversations ]
>> This [inaudible] has an unusual [inaudible] trainer.
[Inaudible] by a robot named [inaudible].
It's one of many iconic aids the staff uses,
including [inaudible] to help care for it's elderly patients.
[Inaudible] reports, this [inaudible] in Japan.
>> Japan's population is rapidly aging
and it's workforce is shrinking.
So it's facing this new shortage of nursing care workers to care
for a growing number of elderly.
And it's planning to turn to robots to help fill this gap.
>> Following [inaudible] Japan's population [inaudible].
There's been decades of [inaudible] robots
as friendly and useful.
People seem [inaudible] they think these are the human touch.
>> And science has made a robotic bed
that splits in half and half.
Half of it turns [inaudible] wheelchair.
Cyberdyne has made an exoskeleton
that you can wear attached to your lumbar area
that helps workers lift like people from their beds.
And probably the cutest robot is a robotic seal called
[inaudible], which is used in [inaudible] and he can respond
to touch, speech, light.
>> Tokyo sees this as a big business opportunity,
and companies have joined them in investing million
of dollars in robot R&D.
But they're also eyeing potentially a bigger prize,
sending these homegrown robots overseas,
especially to other nations with populations [inaudible].
>> Jean-Pierre Auffret: So as you can see, a lot of potential
and possibility for robots and technologies with seniors,
but also a range of issues in regard to like innovation
and scaling and adoption [inaudible] --
So as you could see, a lot of potential and possibility
for robots and technologies with seniors, but also a range
of issues in regard to innovation
and scaling and adoption.
And we'll cover some of those during the course of the talk.
Aging isn't just limited to Asia Pacific.
It's a worldwide phenomena, aging societies.
So many of the European countries
across Asia Pacific including Japan, Korea, Singapore, Taiwan,
China itself because of some of their policies over the years,
some of the South American countries like Argentina
and Chile, and some of the Middle Eastern countries, too,
and North African countries such as Algeria and Iran,
all aging quite rapidly.
And over the coming years going to be approaching 25-30%
or even 40% in Japan's case,
of the population being 65 years and older.
So significance aside all challenge,
and for many countries the risk is
of becoming older before they become developed or richer.
So for a country like Japan or Korea or some
of the European ones which are very successful economically,
have public health infrastructure in place,
their much in better circumstances for aging society
than a developing country that doesn't have public health
yet or, indeed, has a relatively low GDP as some
of the developing countries do.
We'll see in a little bit, too, it's also a phenomenon in Africa
where some of the African countries projected not
to have 40% of their population, but up to 20%
of their population age 65 and older by the year 2050.
So aging oftentimes is depicted as homogenous, or one view,
of what aging is or what aging means.
But, really, aging is quite varied.
Many seniors now and over the years quite active well
into their 60s, 70s, 80s, and 90s --
active in society, active working.
Other seniors, more and more, are trying to age in place
and there's different policies around the world for that --
to be able to stay living in their homes longer
as a way to stay engaged.
But also too many seniors do have fragility issues
and then some also have memory issues.
So very broad range of seniors and a very broad range
of technologies which are applicable to each of the groups
or circumstances of the seniors.
And we'll also see that, how the technologies are being tailored,
including fostered and promoted by the U.S. Government
for different types or circumstances of seniors.
One thing which is a major challenge worldwide,
and we'll talk about some of the other challenges of aging
in addition to health, is this question of poverty.
Seniors oftentimes, and in many countries, don't have pensions
or something has happened to the pension
such as municipal bankruptcy,
or a country doesn't have a pension system yet,
and when the senior's no longer working then then don't have a
regular income.
One other thing which is happening
in many countries is a change of culture.
So many countries traditionally have had a cultural support
system where the young care for and support the parents
and the grandparents and the elderly.
That is changing in many countries now,
including in Asia Pacific where sometimes the youth moves
to the larger cities, but the seniors remain
in the rural areas, and there's not the contact or support
that that might have been in previous decades.
One other issue, too, is with the rapidly aging society,
instead of having six or seven
or eight younger people supporting the seniors
through pensions or care, in some countries it's moving
to three to one or two to one,
two people in the workforce supporting one senior person.
So the support system may or may not be there
from country to country.
So aging varies across the world, and aging, too,
varies across the U.S. but by county and city.
Some counties and cities much more older than others.
On this slide here, darker blue is a higher percentage
of a county with a population of age 65 and over including
with a disability, so having both.
You can see many counties have a bluer shade
and are older than others.
Oftentimes these are in rural areas.
I grew up in Florida which has an interesting
population dynamic.
It's both a very large senior population but also quite young
and somewhat limited in the middle over the years.
So you can see some of these counties quite rural,
quite limited incomes.
With some of the healthcare policy we've had or some
of the circumstances with those rural hospitals closing
and support systems not being what they might be, and so here,
too, in the United States, aging can be a challenging issue.
One other note on disparities in health.
There's the -- have many of you heard about or read
about the stroke belt?
Also a major public health issue.
So there's a band across northern North Carolina,
southern Virginia going across into Kentucky,
and the population there has incidents of stroke
at a much earlier age and much higher preponderance
than across the rest of the country.
So it's a little bit unclear as to the causes of that.
Some mix of probably genetics and diet.
But there, too, there's a disparity of incidents.
And we'll come back to that when we talk
about fragility for seniors.
Here's another chart.
This one is population over 65 by county.
So you see some of the very popular retirement areas
in Palm Springs, also Florida,
but also quite varied across the country.
Oftentimes younger working-age move
to where economies are more dynamic,
but true around the world.
But here, too, in the U.S. Do quite a lot with APEC.
APEC, Asia-Pacific Economic Cooperative,
a multilateral organization of 21 or 23 countries
around the Asia Pacific Region, and here, too,
you can see the vast differences in percentage of aging.
And also vast differences in type of economy and level
of development of economy.
So quite striking the projection for year 2050,
the right-hand column.
Japan projected 42.5% of the population 65 and older.
Korea, which has had the lowest birthrate
for many years now, 41.5%.
Really quite interesting.
Thailand, a middle income country, at 37.1.
So it really varies across the world.
So impact on society and social inclusion, productivity,
all those types of things related to demographics.
These also quite striking type of maps --
this is year 2000 on the upper map,
darker brown the code is more aged like 25 or 30%
or over, 65 and older.
But the bottom map, world map, is 2050.
And you can see that the shift in colors over the years.
Not very many countries would [inaudible] dark browns.
If you can bear with me in 2050 [inaudible] not others.
Oh, yes. Thank you.
[Inaudible] gave a talk last year in Argentina.
It was partially related to productivity in the workforce.
So Argentina, a middle income country,
although with a somewhat sad political history and sadly back
in the news again recently.
But here, too, you can see that the shift in population
and the increasing number of elderly, decreasing size
or stable size of the workforce.
So a challenge for Argentina to increase productivity
to help address their debt issues to be as dynamic
as they could be with their economy.
And think this is the most --
really quite amazing chart in regards to demographics.
So China with the highest population in the world,
but not for very much longer.
So quite rapidly aging with the one child policy over the years.
Relatively soon and into the 2040s
and 2050s the India workforce going to be much, much,
much larger than China.
So great ramifications for China in regard
to competitiveness and productivity.
I didn't bring any charts on this, but China,
the most rapid adopter
in manufacturing, one of, of robotics.
And partially it's to remain competitive.
Partially, too, it's with -- in regard to their aging issue.
[ Inaudible Conversations ]
Okay. One of the challenges I mentioned is the this issue
of workforce and productivity, and we'll come back
to health in a second.
So the Korean population somewhat flat
and will be declining a little bit over the coming years
as their population ages.
Their productivity as a country has been flat or decreasing.
They're of a size where it's quite challenging economically,
so they're not as large
in population-wise as a U.S. or a Japan.
They have about 50 million people.
So for Korea to maintain competitiveness in the face
of a flat or declining population is a great challenge
for them.
Also, too, they have structural limitations from policies.
So the retirement age is quite low in Korea.
They're working to develop policies and approaches to --
so people can stay in the workforce.
Traditionally what's happened is someone will retire
from a very skilled or skill-oriented job
and not really have anything to move to that's comparable,
and might end up doing something
that is not taking advantage of their talents.
So they're trying to address that.
They're also trying to address some
of the employment challenges that they have.
They're dominated by the very large conglomerates
such as the one Samsung is in.
Limited opportunity for youth.
And so there's an office right here on Gallows Road.
It's related to the Korea U.S. Science Technology Center.
But what it does is it enables Korean companies the opportunity
to come here to the U.S. to further their exports or look
to see if the U.S. is a good market for them.
And then the Korean government also has 14 offices
around the world for recent Korean university graduates
to go overseas to get a couple of years of university --
work experience before coming back to Japan.
So Korea, in that middle size, and it's going
to be quite a challenge as that they -- the society ages.
Pensions, too, can be a great challenge worldwide.
So some countries don't have pension systems developed yet.
Others have quite robust pension systems that are under-funded.
This was a news headline from Stockton, California,
a number of years ago when their city went bankrupt.
And so there was a risk to both the pensions
of their former municipal employees but also
that the healthcare benefit
that the city was purchasing for retirees.
So pensions can be a big, big, big challenge
and a big issue both for developed and developing.
A couple of other notes on major trends before we focus more
on the technology in healthcare that are related to aging.
One is urbanization.
Many of the young move to the outskirts of the larger cities
in developing countries, the larger mega-cities.
Lagos, for example, has grown from about 500,000 population
to about 20 million over the course
of the last several decades, so really rapid growth
in a country like Nigeria.
But many, many cities around the world.
Countries struggle to provide city services, provide jobs.
It's also -- help address social inclusion issues.
And so with aging, it's just adds to the challenge
or compounds the challenge.
One interesting phenomena in Japan, very recently.
So Professor Obio [assumed spelling],
a colleague of Yuki's.
Over many years, the youth were moving to the cities
and the elderly were staying in the rural areas.
Now more and more of the youth are looking to the rural areas
for their livelihoods, and there's been a little bit
of a pendulum shift in Japan but with, I think, a lot of benefits
for a more heterogeneous society in the rural areas.
Here's a slide showing the recent change
of the preponderance of population living in cities.
So, of course, here, it's been many, many decades since --
the crossing point is when the more
than 50% were living in cities.
So many, many decades here and in Argentina and Europe.
But worldwide, only fairly recently
that that the cities have become the main hub of livelihood
in many, many countries.
And this also really kind of interesting
from a map standpoint.
This is 1970.
Darker color brown is more higher percentage urban.
And then this is 2030.
You can see many, many more countries have a higher
percentage of urban living,
but also the red circles are larger cities,
like over 10 million in size.
So big shift in how people live and related to aging.
And then other big shift related to aging is chronic illness.
So also only recently was a shift --
that from worldwide mortality go
from the preponderance being communicable diseases
to the preponderance being chronic illness.
And so with aging -- with seniors living longer,
and with improvements in healthcare, more and more
of a challenge of chronic illness.
But also this is true worldwide, not just with seniors.
With chronic illness, it's a much different type
of healthcare than with communicable disease,
much more focus on continuity of fit care,
engaging of allied health, role of diet, role of adherence
to medical recommendations, and a much more important role
of health literacy on the part of the population.
To show that it's quite international, India,
for example, has a very high like occurrence of diabetes,
one of the highest in the world.
And interestingly enough, the higher occurrence in the world
of Hepatitis C is Egypt.
And it came about from a public health failure.
The Egyptian Government was --
they had a parasitic Nile River disease, and they were trying
to inoculate the population against this in the '60s,
but they were using dirty needles.
And so now they have a really major Hepatitis C incidence
in Egypt.
But there, too, chronic illness being a worldwide issue.
And here -- this was a diabetes.
And you can see darker green is higher incidents of diabetes.
So Mexico, Libya, Saudi, Arabia, India, fairly high.
Quite varied around the world of that.
Okay. So what's happening in technology that's really a plus
or an opportunity for healthcare and for seniors
and for staying connected and for living, aging in place,
and living longer, and more engaged in active lives.
A lot of it has to do with the mobile phone and with sensors,
as with a lot in our society.
So the advances have been quite remarkable,
and the opportunity quite remarkable, too,
for really connecting people with services
that they didn't have access to before.
So just in 2015, already 7 billion mobile phone
subscribers, more than one for every person on the planet.
And it's not just -- it's worldwide.
It's developing countries and developed.
A little bit less so in rural areas,
and a little bit less so in some countries.
But these mobile phones with the power
of computers providing services
that weren't really forecastable 10 or 15 or 20 years.
Also, too, this -- the more and more increasing number
of connected devices and sensors as we'll see
for assisted technology really providing capability
that wasn't there before in the whole mobile app inroads.
[ Inaudible Speaker ]
One kind of note which I think is neat in regard
to how rapidly technology changes and how it's difficult
to forecast where it may be in 10 or 20 years --
these are just pictures, but -- of status storage,
transmission speed, and processing power.
But just to show how far we've come, the first one --
so say you can go into a computer store
and buy a several terabyte storage like near
where I live, where we live.
It's like 30 or $40 or $50 for like three terabytes,
and it's the size of a book.
The first commercially available gigabyte storage was just 1980,
so not that long ago in many ways, and it was 550 pounds,
the size of a refrigerator, and cost $40,000.
And it was marketed by IBM.
So from this and one gigabyte of $40,000 to about this
for multiple terabytes from $50 has been the change.
And so that change is continuing.
And so the underlying cost of storage, the variable cost,
is going to be almost zero shortly.
The processing speeds similarly.
So now it's six, seven, or eight going in cycles a second
in a computer on research.
Kind of a coincidence --
the first lunar lander and also the first Apple computer were
both 2000 cycles a second.
So from 1970 to today, from 2000 to 8 billion,
and that's advancing, too.
And then the transmission speed -- the first transatlantic cable
from 1957 could only handle about 36 simultaneous calls.
And now it's billions and billions of calls
on a fiberoptic network.
So with those trends continuing,
just as you couldn't really foresee the capability
of the iPhone if you were back in 1990,
you can't really foresee what the computers
or the phones will be doing in 2025.
You have an idea of what the applications might be in regard
to [inaudible] with cars and sensors or with videos
and images and holograms,
but the specific applications you can't really --
it's hard to visualize.
This is a neat resource that comes out every fall from ITU,
a Measuring Information Society.
The blue line that's going
up towards 100% is mobile phone subscriptions,
so about 100 mobile phone subscriptions per 100 people
around the world.
The really dark blue one that's quite steep on the right --
that's mobile broadband subscriptions.
And you can see that's increasing more rapidly
than mobile phone, so probably in several years we'll be
up to 70, 80, 90 mobile broadband subscriptions per 100
population around the world, developed
and developing countries.
So great, great, great change.
And this, too, I think is another background
on what's possible, what technology.
For Mexico's to the left, shows the percentage of the population
in Mexico that is banked or has bank accounts.
The right shows the percentage
of the population with a mobile phone.
So much, much, much higher.
So Mexico, one of the countries where these types of apps,
where you can do banking through the phone like M-Pesa in Kenya,
really making inroads.
So instead of having traditional bank accounts, going to one
of these phone-based bank systems
and really providing access.
And so -- also for health, too.
Of course, the whole smart city playing a big role
in aging society as we'll talk about as well as the smart grid,
so really changes in how electricity is provided.
I didn't include a slide, but Africa, for example,
just in the last 20 years and projected out to 2020 the amount
of power which has been brought online,
and related to the phones, or enabling the phones,
has really led to major shifts.
One other video I want to play before we talk more
about health is not all technology has
to be really leading edge.
And it's possible to do things in health with technology
that is somewhat dated such as SMS.
And there's some wonderful examples
around the world including here in the U.S.
of using SMS-type technology to really further health outcomes
and benefit seniors in a way just as much as some
of the more advanced robotic or AI technologies.
So this is a really neat initiative in Tanzania
with Novartis and IBM.
[ Speaking in Foreign Language ]
[ Music ]
So the idea is that the mortality
from malaria is very high for children under the age
of five without treatment.
And Tanzania there having very, very high incidence
of malarial medicine stock-outs in rural clinics.
And so with Novartis and IBM
and a telecom [inaudible] developed a system
where the clinician would just send a text message
in once a week on her inventory level.
And if the inventory was low or nearing a stock-out,
it would be replenished.
And they went from 80% or so stock-outs to a very,
very small number with the corresponding benefit
for the children and for children's health.
There's a number of amazing examples similar to this.
Pharmaceutical counterfeiting is rampant in some countries.
And an almost similar system in one or two
of the West African countries is you don't buy the medicines
from a pharmacy like we do here, like a CVS.
It's more from a small shop.
But just a code on the side of the medicine that you can type
into your -- send a text message with your phone
and it validates it that it's real.
So here we have a new law that's going to take effect early 2020s
on a pharmaceutical traceability.
And the -- what the legislation calls for is the ability
to trace a pharmaceutical all the way from manufacturing plant
to when we actually pick it up at the pharmacy.
And so much more advanced than this text message idea.
But some of the pharmaceutical companies are looking
at block chain for that, for example.
Okay. So technology doesn't need to be advanced, but, of course,
a lot of the technologies are advanced including with robotics
with a lot of promise for seniors but a lot of promise
for society and a lot of risk also.
The Asian Development Bank puts out an annual report
in the spring, and it just came out this year.
And they have a fascinating chapter on the job impact
in developing Asia countries from robotics.
It was quite a, I thought, sophisticated study
because they had the downside risk industry by industry,
from country to country.
They also had the upside potential
of access to the job markets.
And then they also looked at a forecasting
when robotics would be adopted industry
by industry based upon the cost, the economic opportunity as well
as the technological capability.
But many robots used for different applications.
The one on the bottom right is an English-speaking robot
in Korea which is really a popular teacher.
The upper left is the more traditional view
of industrial robots.
But you can see by comparing the two,
that the robots are quite different.
So robots are becoming more non-tethered.
But the one on the left is tethered and stationary.
More mobile and non-tethered, more autonomous
and more interactive and in some ways more
"intelligent," in quotes.
So through all of that, much greater interaction
with humans and individuals.
And so many more risks for the individual.
Many more societal implications and public policy implications
as these larger and faster robots are interacting
on a day-to-day basis.
So there's a lot of studies
which are really quite interesting on the use of robots
and humans together as to whether
or not a human plus a robot has greater capability than a human
on its own or a robot on its own.
In one of the studies which gets quoted a lot is a chess study.
So pairing -- this is from several years ago --
pairing the chess-playing programs that the real top ones
and the not quite top ones -- with chess-playing humans,
really, really top ones and not so top ones.
And the winning teams from the experiment were --
it wasn't that the best human or the best robot or AI.
It was the one that worked best together.
And you can see that's going to be a factor in many fields.
So, for example, medical diagnosis --
the IBM Watson does commercials quite --
or commercials about Watson quite often.
Doctors are using Watson at Cleveland Clinic I believe.
But there it's an interaction of what the AI tells the doctor
and what the doctor interprets and what the --
what information that the Artificial Intelligence provides
to the doctor so the doctor can make a judgment
over how accurate it is.
One kind of aside on that.
It's a major, major issue that this question
of artificial intelligence and transparency.
Here in the U.S., but also in Europe,
and there's a really interesting case with Zillow.
Does anyone follow that case?
So many of you probably use Zillow or seen it.
So Zillow is a home and condominium site.
And one of the things it has on there in order
to draw traffic is it has an algorithm
that generates what a house is worth or valued or would go
for on the open market today.
And so you can go on there any day and type in a property
or your own address and see what it's valued at.
It also has a chart for historical estimates as well
as for the neighborhood.
And so an attorney saw that her house was being undervalued
on Zillow, a very, very nice house, high-priced house,
and it was hindering her ability to sell it.
And so she sued Zillow and -- but Zillow didn't --
and so the judge wanted to see how Zillow operated
and how Zillow came up with the estimate
of what the house was worth.
But Zillow said it was a trade secret, how it was doing it.
So the judge had to make arrangements to hire an expert,
but the plaintiffs paid, to review the system
in order to make a judgment.
So that works in the cases when it's an algorithm or can work,
but in these issues --
circumstances with machine learning that the AR
or the machine gathers information from the Internet
and then slowly over time develops an insight and a view.
It's much harder, if not impossible,
to know how the machine came
up with its recommendation or solution.
But the laws now in Europe -- there's a transparency law
and algorithms that you have to be able
to show how a computer arrived at a decision.
And also, too, that there's laws so that someone can ask
for a human to make a choice instead of a machine.
So as an example, a university hypothetically could use a
computer to decide who's admitted, really quite possible.
And some jurisdictions have experimented with computers
as to whether or not someone should be paroled or whether
or not they'll be at risk if they're paroled.
So the -- a lot of the policy orientation is what recourse
does an individual have because you can't really depose the
computer necessarily, and then so what recourse is,
and how do you know how it made its choice?
And then what can you really do about it?
Or even, for example,
if a autonomous vehicle caused an accident when you go --
and if someone was going to sue the insurance company,
you can't really depose it, or how do you find out about it?
This is actually in interesting picture, too.
So this is an assisted robot for Japan.
So with a rapidly aging society, limitations on the number
of health workers to assist -- so this one to help lift someone
out of a bed and move from one bed to another
or when a bed is being made.
You can imagine the legal risk on something like that.
As it turns out in the U.S. that there are --
every year there's an incidence of people being,
patients being dropped in hospitals, too.
So -- but that one being adopted in Japan.
And I think we would see something
like that here over time also.
One of the arguments from the autonomous vehicle providers is
that the autonomous vehicle will be safer than a human driver.
But that's not the view of all the car companies.
Toyota, for example, their view is more
that the autonomy is assistive.
And so there's a range of circumstances where visibility
and rain, when the autonomy is really helpful.
But not at all times necessarily.
This is a very famous Japanese robot,
Chihira Kanae, from Toshiba.
One thing to note is Japan's going
to be really highlighting robots
for the 2020 Olympics, including at Narita.
But there's different cultural views
of what a robot should look like.
So, in Japan they have the video reference Paro for the autism
and aging, a very cute and attractive seal.
This robot looks much, much different.
So this one looks as if a human.
So there's cultural differences on that.
So I think in the U.S., this is a broad generalization,
but for the most part, our robots are idealized
to some extent, and less human looking.
But some countries, they do have robots
that are a little bit more human looking.
So, again, cultural differences in robots.
Here's a initiative here from the U.S., from HHS.
Research -- I think this was from last year,
the announcement request for proposal for robotics
with Alzheimer's assistance.
So many, many opportunities for research grants in --
at the end here, I'll discuss a little bit how our government
tries to catalyze research.
I have to be very cognizant of the time.
Two other areas which the great [inaudible] impact for seniors,
nonhealthcare, is one is autonomous vehicles.
This is actually a drone in Rwanda.
So Rwanda's very hilly.
And so this commercial company to deliver fixed-wing drone,
deliver medicines to rural areas
because the roads aren't necessarily that good.
And they do deliveries within two or three hours.
They can get their order in and ship it out.
But the point more in general is that autonomous vehicles,
especially for seniors living in rural areas
where they've lost the bus routes,
which happens all over the world.
Now with autonomous vehicles there's a way
where the senior could actually have a vehicle
that can bring them into town or they'd be much more mobile
and more able to live and age in place.
This is just kind of -- this is a autonomous drone in Dubai.
I don't think I would get into one like that, but that's there.
And then, too, the artificial intelligence becoming much
more flexible.
So chess, although very complex and complicated,
somewhat limited in regard to its one application.
But more and more some of the AI applications are quite flexible
as to what they're doing
and quite large ramifications, I think, for society.
We use AI every day, most every one of us, in some ways.
So, for example, someone going shopping on Amazon,
and the recommendations that come from Amazon,
sometimes they're a little bit creepy I think,
the types of things that they do.
But a lot of that is from AI and gathering information
about a person and then utilizing it in a way
to [inaudible] back a recommendation.
There's a projection for Philippines
that in the next two years they're going to lose 30%
of their call center employees.
So Philippines, a major industry with call centers.
So you call from here and you get someone
from Philippines quite often.
So a lot of the first and second level that the levels
with not the huge expertise, the projection is they're going
to be replaced by automated assistants in the next years.
So a very, very big issue for Philippines
with job replacement and AI.
And then the last thing [inaudible] trend, I think,
is this open data issue that the data being available
and the potential both for benefit for seniors but also
for society in building businesses.
So here in our area we have a quite robust open data,
the federal government obviously,
the counties, New York, too.
Some of the really neat applications I think is what you
can do with trending over time for seniors.
There's a really neat company here locally.
It's now possible to get from --
a [inaudible] from HHS much more quickly than years ago,
information on Medicare and Medicaid payments and care.
And it's six months or probably less now.
And so local company here obtains the information
from HHS, then they're able to do comparisons from region
to region or from hospital system to hospital system
and provide the information back out to enable the caregivers
to hopefully provide better care.
Okay. So how is technology actually being used for aging?
And it really varies by the type of --
where the senior is in aging.
So again in place, an active aging type
of technology may well be in exercise
or an autonomous vehicle.
Someone who's a little bit more fragile might be assisted
for mobility or for sight or with chronic illness.
Then, too, for memory care,
such as with the Paro from Japan, the seal.
So aging very, very, very different.
And then also the government services vary too.
The one on the right from New Jersey I think,
a really important type of initiative in times of emergency
to know where the seniors are, to know where the ones
who have memory issues are so you're able to better evacuate.
So it's a huge impact and benefit both
to the senior themselves and to the city if you know
where everyone in your city is who needs an oxygen tank
in regard -- if the power goes off.
So those types of things are big, big benefit.
One of the things I think that --
really quite important in aging is this concept of fragility.
There's probably 80 different models of how to define it now.
But fragility both of movement, but also fragility of health --
inability to take a number of stresses to a person's system.
So someone very young in their 20s,
they can take multiple health stresses and be fine.
But someone older is much, much more difficult.
Fascinating study from Japan on seniors.
And this is with fragility in movement.
It's very, very -- their study found that it's very difficult
if you move from quite active to limited mobility and fragile
in movement, and then almost non-mobility to go back,
to make your way back from very limited --
almost no mobility to fragile and fragile to active.
So a lot of the interventions are to --
if someone is to prevent someone or help someone
so they don't move to the next stage.
And there's also models like this for cognitive fragility
and the intersections of the two,
more physical and more confident.
But a lot of the technologies are oriented towards this type
of idea that to help someone stay in an earlier stage,
a more healthy stage, and help them move back
or to the extent it's possible as they move further.
Because everyone knows someone who's gone ill when they're
in their 70s or 80s, and it's much more difficult recovery,
much more challenging or a fall is much more challenging.
This is a similar model for WHO on functioning and disability.
It's quite a similar idea, but here it's the degree
of societal participation and then functioning, physical
and mental functioning in the intersection of the two.
So the point being there's a great benefit
from engaging societally or with family and a risk
from social inclusion.
And so -- and also a benefit from some physical activity.
So what are some of the technologies?
This is really a quite neat one I think.
And this is an [inaudible].
Canada has, and Taiwan and New Zealand is actually a leader
in this, have implemented these fairly widely in some cities.
So enabling the phone or the cane
with communications technologies to connect
to beacons or sensors.
So a lot of the sight-impaired now have GPS on the phone
that can direct them to a place where there's GPS.
But if in a very large building or a subway station,
they don't have the GPS.
But to put beacons in there and to communicate back
with the cane or the phone or the eyeglasses as in Canada,
and really help someone there with their mobility
and their confidence to get around.
And it can be really very, very lifechanging.
There's a video, which I'm not going to show,
of someone about 30 years old who -- in Canada --
who lost their sight through diabetes.
And she now has these types of technologies and is really able
to get around and do quite well.
So a wonderful lady at NIH who also --
who lost her sight when she was about 25
and she afterwards moved --
started in information technology and, again,
is able to do tremendously with the assistance of technologies.
This is a second one, cognitive ability,
which a lot of technology's very helpful with.
Both continuing with -- to be engaged, but also some
of the technologies more and more able to identify trends
and patterns for early intervention.
For example, with dementia or Alzheimer's.
So I think with a lot that's being done nowadays,
a lot of potential.
One thing that's kind of interesting.
A George Mason professor I work with has worked on this.
The way someone uses their Smartphone is very unique.
You can identify someone from the Smartphone usage.
So, of course, the intelligence agencies are very interested
in that sort of thing.
But also, too, there's a thought
that over time someone might have a change
in Smartphone pattern usage,
how they actually type on the Smartphone.
And you can identify insipient dementia from that.
So there's a number of projects like that both here
and the U.S. but around the world.
This is a DARPA Drive technology.
Really tremendous potential for seniors
for mobility and staying active.
This is an exoskeleton.
But many, many different ways of doing this.
There's a number of technologies that haven't worked
yet that I think are really interesting if they could work.
Types of technologies that a senior could wear, very light,
that would expand, for example, on a fall to make it less likely
that a fall would be catastrophic.
You're tied with sensors so you might have an iPhone.
This is a local company here in Fairfax
which is really kind of remarkable.
So they have this -- the EyeGaze Technology has come a long way.
At Consumer Electronics Show they had an exhibit,
a company on exhibit this year where you could drive a car just
by moving your eyes and steer it.
This company here -- this is
for someone who's had a really catastrophic accident
and is paralyzed from neck down -- that they can --
someone can actually type and write by moving their eyes
on this keyboard on this screen.
And when they come -- and we've had several assisted technology
conferences, and the company comes and gives a demo,
and some of their customers come with stacks of books
that they've written using this technology.
So it's for someone in that sad circumstance.
It's really lifesaving.
It gives them a real path to life I think, in livelihood.
A lot of the senior technology --
this is for [inaudible] you know, profession soccer players,
but we [inaudible] this as senior technology with sensors.
So this technology here, which is being adapted,
is having a communications device in the soccer ball
and having sensors around the stadium, and then also
on each player having a communications device.
So you're able to track the ball
and all the players in real time.
And then make judgments about how they're playing
and how often they have to have the ball.
So their golf technology --
if you ever watch gold on T.V. they have --
they can actually track the speed of the ball off the club,
and they do it with baseball now, too, and follow the path.
But this having a big impact.
And then lastly, too, on chronic disease management,
this is a Google initiative for diabetes.
So being able to tell insulin levels from tears
with a contact lens, communicating either to a phone
or to a SmartGlass, and then being able to provide a warning
if the insulin levels are --
or the patients are outside of some sort of control
and that's -- so better care and more stability of blood levels.
So, again, a neat thing that technology's making possible.
So to wrap up, what are some of the major challenges
of getting these types of technologies to seniors
and to patients in general?
What is the phone coverage and digital divide?
So even though there's almost one phone per worldwide
inhabitant, in some areas the phone coverage can be a little
bit iffy, the broadband coverage.
And so that can be a major, major issue.
Health literacy could be a major issue and challenge, too.
I think this goes both ways though.
So there's -- health literacy is important
for using the technology but the technology can help
with health literacy.
There's a Georgetown company right there by the canal
and they were -- they provide
through Smartphones text messages
to low income expectant mothers, about-to-be mothers.
And their service was highlighted by President Obama
and they're in many hospitals nowadays with this,
and they've had -- they were able to demonstrate a benefit
for both maternal health and infant health.
So the messages are coming from pediatricians,
so they're real healthcare providers.
But they're able to demonstrate that type of benefit
through the technology, and it's measurable.
Doctor engagement also, too.
This varies by culture.
So in many or most countries' cultures there's a hierarchy
that a doctor, being quite --
very, very well respected, obviously.
But the technology changes the dynamic a little bit.
And it changes it differently country by country.
So in some countries the patient's very empowered
and the doctors don't necessarily like that.
In some other instances, the technology
and the information the patient has may
or may may not be all that helpful.
One thing that's very much
in the news is electronic medical records
in patient reporting,
and to what degree patient reporting is helpful
to a doctor.
And some of the better electronic medical records flag
the provenance of the data as to whether it's
from a real healthcare provider or from the patient.
And one also interesting aspect of that is there's more
and more consumer technology that people wear
like Pit Bitz and things.
Many, many, which are quite high quality.
So the information is 7 by 24 as opposed to clinical trials
where you go to the hospital once a week and you're tested.
And so there's a couple of clinical trials --
one at Duke and one at Stanford --
to use commercially-available technology with 7
by 24 information and then seeing
if you can make a judgement on a new other device or treatment.
Some of the other issues -- I'll skip some.
Obviously payer models.
Reimbursement really, really important
for any technology to be adopted.
Here in the U.S., our laws changed last year.
So, historically, for telemedicine --
for telemedicine to be reimbursed you have to be
in a medically underserved area
which was most often a rural area, or a city, a very,
very urban inner-city-type area.
But it changed last year, so it's much more reimbursable now
to -- - - for -- to -- back to the insurers.
And also more -- Medicare and Medicaid more open to paying.
Standards, obviously, interoperability.
Market size has to have a financial benefit
to private sector companies.
And then adoptability.
Some companies and some research projects are fantastic
at engaging seniors and engaging the patients in design.
Others not so.
I think that's true with IT in general.
Some of the IT development projects are fantastic
in engaging users in advance.
And then others not so much oriented that way,
and the system is finished and might not be all
that fit for a purpose.
So engaging the users is very important.
I'll wrap up in a second.
This dual use idea is really quite interesting, too.
A lot of commercial -- military technologies are being
repurposed for aging and for assistive needs.
Some of the hand motion technologies being adapted
for autism and autistic children.
Another need autism technology -- this is local also.
For a severely autistic child they have 20 or 30 caregivers,
so they have -- well, the parents, obviously.
And then they have psychologists,
they have language therapists,
they have their teachers in the schools.
So many physical therapists and many, many, many.
And as they go through the course of the day,
periodically a child like that might have a emotional incident
or outburst.
But they're somewhat predictable and the technology
with this company here locally --
they try to connect the caregivers in the chain,
and if an early-morning caregiver sees a sign,
they'll flag it to the next ones so that they can react to --
interact with the child differently based upon how the
child's day is going.
So again, just phone technology but kind of neat.
And lastly, policy is very, very important.
Most countries now have national ICT policies.
And then most have also ageing policies.
It's the rare country where there's an intersection
of the two, where there's a connectedness
of the ageing and ICT.
But tremendous potential.
Japan is connected there.
Theirs is quite well.
But others are working on it.
And then one last note here in the U.S. I think one of the --
two of the best things we do from a technology standpoint.
One is the small business innovation research grant
program which was the Alzheimer's
and robotics grant was under that,
for fostering commercialization of technologies
through small businesses.
It's a world leader.
Many countries are trying to adapt it
to varying degrees of success.
But a small company with an idea or somebody with a prototype
in an area of interest to the government, there's quite a bit
of funding both to develop it to a trial and then
to actually commercialize it.
And another good, really good thing we do, I think,
from a research standpoint is DARPA.
We do many good things.
But DARPA has been very successful.
So a lot of the autonomous vehicles technologies today
and around the world and here, many companies working on it.
It started at DARPA only about 12 years ago
with a DARPA Grand Challenge for teams to come
and develop an autonomous vehicle that was able to drive
through the Mojave Desert for 108 miles along cliffs
and over hills and everything like that.
In the first year, the winning team [inaudible] went
eight miles.
And then all of them had crashed by then.
And so, just a number of years later,
this technology really being adapted worldwide.
And so the DARPA idea was two-fold.
But they had a lot, there were a lot of initiatives
on the different elements of a car.
But there weren't initiatives to integrate them all together.
And that the autonomous vehicles back then weren't military grade
and couldn't go 35 or 40 miles an hour.
And their view was, there weren't enough research teams
or experts in the U.S. and by developing this challenge,
that they foster interest in this
and greatly increase the number of research teams.
And so, in more recent DARPA challenges has been robotics the
last several years.
And so the first one, 2013 and '15, the robots were --
they bounded it for robots
in natural disasters like nuclear relief.
But the first year that the robots were very limited
and couldn't go upstairs or couldn't really do much.
And just in a few years,
a lot of the technology rapidly advancing and cost is going
to be coming down and much more available to people.
So with that, thank you very much for coming.
It's a fascinating area and an ageing society has many,
many needs and especially with the intersection
of ageing society and disability.
So thank you very much.
[ Applause ]
>> Sandra Charles: Thank you for this.
>> Jean-Pierre Auffret: Oh, thank you.
>> Sandra Charles: That was most fascinating and informative.
>> Jean-Pierre Auffret: Oh, thank you.
>> Sandra Charles: Are there questions for Dr. Auffret?
>> My name is [Inaudible].
Thank you for [inaudible] subject wide-ranging,
still very concentrated, interesting as well
as educational insight on the subject.
Your presentation, very [inaudible] to look
at several other aspects.
One of it then is entrepreneurship.
How entrepreneurship could provide [inaudible]
in this sense.
We, my group [inaudible] work at the sound [inaudible] last year.
We were [inaudible].
With regards to that, look
at how the [inaudible] could be entrepreneurial.
It could be not [inaudible] more opportunities.
I just wonder how you look at a [inaudible] rather than kind
of the receiver of the technology as a senior citizen.
>> Jean-Pierre Auffret: Well,
I think it's a very neat question because I think of --
>> Sandra Charles: Could you [inaudible]?
>> Jean-Pierre Auffret: Oh, but the -- thank you.
[Inaudible] that the question is about the potential
for senior entrepreneurs developing technologies
in addition to just being users.
I think with increasing life expectancies
that many seniors retire at 60 or 65 or 70
and will live to 90 or 95.
So they have years for a full career ahead of them.
And those that are able to remain active and vibrant
and healthy, really would also benefit from an avenue or a path
and have a tremendous amount to offer
and a tremendous amount of insight.
So I think that there's many senior entrepreneurs and I think
that the potential is really great.
And they also would bring insight
to these types of technologies.
But that the pensions too, I think,
in social security also probably to some extent,
what was designed when life expectancies were much less
than today.
And so with seniors living, population living so much
to much greater ages, potential for second and third careers.
On one other note, I think it does relate to the whole idea
of careers is somewhat different now.
Much less of it in the U.S. but many countries too,
where someone works for one company for many, many years
and also with the way technology changes so quickly.
Industries seem to come and go also.
And so most people, through the course
of their lives even before 65, have three or four careers.
So a really excellent question.
>> Sandra Charles: Any others?
>> Jean-Pierre Auffret: Okay.
Well, thank you all for coming.
>> Sandra Charles: Thank you very much again [applause].
>> This has been a presentation of the Library of Congress.
Visit us at loc.gov.
Không có nhận xét nào:
Đăng nhận xét