WEBVTT

00:00.469 --> 00:03.495
>> Emerging technologies.

00:03.495 --> 00:06.180
The learning objectives
for this lesson are

00:06.180 --> 00:08.490
to describe blockchain concepts,

00:08.490 --> 00:12.120
to explain artificial intelligence
and augmented reality,

00:12.120 --> 00:13.635
and to discover big data,

00:13.635 --> 00:16.080
deep learning, and
quantum computing.

00:16.080 --> 00:20.685
Let's get started.
Blockchain. You're probably

00:20.685 --> 00:21.900
are familiar with this at

00:21.900 --> 00:24.180
some level because
you've heard of Bitcoin.

00:24.180 --> 00:27.210
Bitcoin is the most
well-known example

00:27.210 --> 00:28.635
of the blockchain in use.

00:28.635 --> 00:30.405
But the blockchain
can be used for

00:30.405 --> 00:33.915
many different types
of problem-solving.

00:33.915 --> 00:37.500
The blockchain at its basic
level is an expanding list of

00:37.500 --> 00:41.489
transactional records that
are secured by cryptography.

00:41.489 --> 00:44.945
Each record is known as a block,

00:44.945 --> 00:47.615
and then they're all connected
together in the chain.

00:47.615 --> 00:50.270
Each block is hashed and

00:50.270 --> 00:53.125
the hash value of the previous
block is included with it.

00:53.125 --> 00:54.785
This way, each block

00:54.785 --> 00:56.960
validates the next
block in the chain.

00:56.960 --> 00:59.465
One thing to keep in mind
is that the blockchain is

00:59.465 --> 01:00.620
a public ledger that is

01:00.620 --> 01:03.110
distributed across
peer-to-peer networks.

01:03.110 --> 01:05.300
Back to Bitcoin, a lot of people

01:05.300 --> 01:07.490
think that Bitcoin is anonymous.

01:07.490 --> 01:11.405
Now we might not know who owns
a specific wallet address,

01:11.405 --> 01:14.870
but every transaction that
wallet has ever done in or

01:14.870 --> 01:19.445
out is available for everyone
to see on the blockchain.

01:19.445 --> 01:22.210
This is something that's
not commonly understood.

01:22.210 --> 01:24.500
But since it is a public ledger,

01:24.500 --> 01:26.825
anyone can view the records.

01:26.825 --> 01:29.300
Those wallets again,
are not anonymous.

01:29.300 --> 01:30.770
We don't know who
owns it necessarily,

01:30.770 --> 01:33.570
but we do know what that
wallet has been doing.

01:34.310 --> 01:37.765
Secure multi-party computation.

01:37.765 --> 01:39.170
This is distributing

01:39.170 --> 01:41.540
computations across
multiple systems.

01:41.540 --> 01:45.050
But the key is that
no individual system

01:45.050 --> 01:46.700
is able to read the data of

01:46.700 --> 01:48.965
the other parties
in this grouping.

01:48.965 --> 01:51.260
This allows us to solve

01:51.260 --> 01:53.479
complex problems
without compromising

01:53.479 --> 01:55.385
the privacy of the data.

01:55.385 --> 01:58.190
Some examples of when we
would use this maybe would be

01:58.190 --> 02:02.940
DNA research or research
into patient data.

02:04.270 --> 02:06.995
Distributed consensus.

02:06.995 --> 02:09.935
A distributed or
decentralized system where

02:09.935 --> 02:11.360
all the systems come to

02:11.360 --> 02:14.270
an agreement for a
specific computation.

02:14.270 --> 02:15.830
We're doing this to maintain

02:15.830 --> 02:19.430
the overall integrity of
a distributed system.

02:19.430 --> 02:21.890
Because we know that some
of the systems that are

02:21.890 --> 02:23.930
included in this
might be malicious,

02:23.930 --> 02:26.465
the distributed consensus takes

02:26.465 --> 02:29.480
a vote from all of the
systems about the data.

02:29.480 --> 02:31.670
The value with the most votes

02:31.670 --> 02:33.810
is the one that gets accepted.

02:34.630 --> 02:37.370
Artificial intelligence.

02:37.370 --> 02:39.110
This is the science of creating

02:39.110 --> 02:40.700
computing systems that can

02:40.700 --> 02:43.040
simulate or demonstrate
intelligence

02:43.040 --> 02:45.170
that's similar to the
level of a human.

02:45.170 --> 02:48.695
Machine learning uses
algorithms to parse data,

02:48.695 --> 02:51.755
then develop strategies
for using that data.

02:51.755 --> 02:55.040
Machine learning can
modify the algorithms and

02:55.040 --> 02:56.570
make gradual improvements in

02:56.570 --> 02:59.040
its ability to make decisions.

03:01.070 --> 03:04.000
Adversarial artificial
intelligence.

03:04.000 --> 03:06.665
Now, whenever a new
technology comes,

03:06.665 --> 03:10.135
it's not long before the bad
guys make use of it as well.

03:10.135 --> 03:12.770
An attacker can gain access to

03:12.770 --> 03:15.590
a network that uses AI
for threat intelligence.

03:15.590 --> 03:17.180
They can inject traffic into

03:17.180 --> 03:20.090
the data stream with the goal
of concealing their tools.

03:20.090 --> 03:22.550
For example, an attack tool

03:22.550 --> 03:24.275
could be disguised
as a text editor,

03:24.275 --> 03:26.000
and then, therefore,
is not a threat.

03:26.000 --> 03:27.530
This requires knowledge of

03:27.530 --> 03:29.240
a particular AI
algorithm that's in

03:29.240 --> 03:31.895
use on the victim system.

03:31.895 --> 03:34.295
In addition, there are a lot of

03:34.295 --> 03:36.290
software packages for

03:36.290 --> 03:38.210
malicious intent that
have been seen on

03:38.210 --> 03:40.430
the market that
allow attackers to

03:40.430 --> 03:45.215
direct artificial intelligence
attacks toward a victim.

03:45.215 --> 03:47.090
These programs are able to make

03:47.090 --> 03:48.950
decisions faster than humans.

03:48.950 --> 03:52.395
When they do get input data
back from say, for example,

03:52.395 --> 03:54.260
an Nmap scan, they can more

03:54.260 --> 03:58.440
quickly automate the attack
back based on those results.

03:59.340 --> 04:02.530
Virtual or augmented reality.

04:02.530 --> 04:05.845
This is an extended concept
of artificial intelligence.

04:05.845 --> 04:08.320
It emulates a
real-life environment

04:08.320 --> 04:11.455
with computer-generated
sights and sounds.

04:11.455 --> 04:13.840
We have numerous
applications for this.

04:13.840 --> 04:15.220
They can be used for training,

04:15.220 --> 04:18.670
providing information to a user

04:18.670 --> 04:20.620
on people or objects
within view.

04:20.620 --> 04:23.200
Imagine you need to
get someone up to

04:23.200 --> 04:25.825
speed pretty quickly on a
given training situation.

04:25.825 --> 04:26.950
A good example of this would be

04:26.950 --> 04:28.555
a military-type situation.

04:28.555 --> 04:31.870
By having a virtual or
augmented reality system,

04:31.870 --> 04:35.090
you can walk someone
through a given scenario or

04:35.090 --> 04:36.670
a given training situation to

04:36.670 --> 04:39.585
help them pick up the
information a lot quicker.

04:39.585 --> 04:42.860
Then on the flip side,
we have marketing as

04:42.860 --> 04:45.380
another issue where you
have to look at with this.

04:45.380 --> 04:47.330
While we're walking
through a given area,

04:47.330 --> 04:48.830
marketing information may pop

04:48.830 --> 04:50.975
up as we pass certain stores.

04:50.975 --> 04:53.120
Or as we look at specific items,

04:53.120 --> 04:55.640
we may see information
on our screen that tells

04:55.640 --> 04:59.220
us more details about
it and maybe a coupon.

05:00.140 --> 05:03.290
Big data. These are

05:03.290 --> 05:04.940
data collections
that are too big for

05:04.940 --> 05:07.220
traditional database
tools to utilize.

05:07.220 --> 05:09.395
These are ideally suited to AI

05:09.395 --> 05:11.540
as the larger the data
set for AI to study,

05:11.540 --> 05:17.055
the more effective it
will be. Deep learning.

05:17.055 --> 05:19.610
This is a type of machine
learning that takes

05:19.610 --> 05:20.720
apart the knowledge and then

05:20.720 --> 05:22.415
breaks it into smaller parts.

05:22.415 --> 05:24.740
Complex topics can
be broken into

05:24.740 --> 05:27.965
parts that are easier for the
deep learning to interpret.

05:27.965 --> 05:31.445
A deep learning system can
then decide which parts are

05:31.445 --> 05:33.185
applicable to a given problem

05:33.185 --> 05:35.800
and make decisions
based upon those.

05:35.800 --> 05:38.450
IMB's Watson is a
good example of this

05:38.450 --> 05:44.370
in action. Quantum computing.

05:44.370 --> 05:47.390
This manipulates data
at the atomic level.

05:47.390 --> 05:49.640
Qubits or quantum bits are

05:49.640 --> 05:52.060
the base unit in
quantum computing.

05:52.060 --> 05:55.955
Qubits have a value or
state of zero or one,

05:55.955 --> 05:58.220
or any value in-between.

05:58.220 --> 06:01.710
They can have multiple
states at the same time.

06:03.860 --> 06:07.565
Qubits can become entangled
and then the value

06:07.565 --> 06:10.985
can only be observed by
collapsing the quantum effect.

06:10.985 --> 06:13.340
The only way to get
measurements is to

06:13.340 --> 06:16.100
indirectly entangle two cubits.

06:16.100 --> 06:18.140
This allows quantum computers to

06:18.140 --> 06:20.615
perform many calculations
at the same time.

06:20.615 --> 06:23.300
Because of this, quantum
computing is especially

06:23.300 --> 06:26.880
useful in breaking RSA
and ECC encryption.

06:27.590 --> 06:29.985
Homomorphic encryption.

06:29.985 --> 06:35.450
These are primarily used to
share privacy-sensitive data.

06:35.450 --> 06:37.670
It allows for
statistical analysis of

06:37.670 --> 06:40.430
data without decrypting
the data itself.

06:40.430 --> 06:42.680
The main purpose
is to share data,

06:42.680 --> 06:43.895
but at the same time,

06:43.895 --> 06:46.200
keeping the data private.

06:49.210 --> 06:51.980
Components of
homomorphic encryption

06:51.980 --> 06:54.995
are private information
retrieval or PIR.

06:54.995 --> 06:57.755
This protocol allows for
the retrieval of the data

06:57.755 --> 07:00.995
without revealing which specific
item is being collected.

07:00.995 --> 07:04.130
We also have secure
two-party computation

07:04.130 --> 07:06.920
or secure function
evaluation, SFE.

07:06.920 --> 07:08.735
This allows two parties to check

07:08.735 --> 07:11.590
input without
disclosing the results.

07:11.590 --> 07:15.785
Finally, we have private
function evaluation or PFE.

07:15.785 --> 07:19.400
This describes
calculations done by

07:19.400 --> 07:20.900
more than one system but

07:20.900 --> 07:24.810
the function used is
only known by one party.

07:26.240 --> 07:30.060
3D printing. 3D printers are

07:30.060 --> 07:31.970
special-purpose
printers that build

07:31.970 --> 07:35.000
3D objects rather than
printing on flat paper.

07:35.000 --> 07:38.465
Printing is done by adding
layers on top of layers

07:38.465 --> 07:39.950
according to a model using

07:39.950 --> 07:42.365
a computer-aided design
or CAD software.

07:42.365 --> 07:44.300
This allows for rapid design and

07:44.300 --> 07:46.385
creation of just about anything.

07:46.385 --> 07:48.710
It has been used in
health care to create

07:48.710 --> 07:50.795
repair parts quickly
and inexpensively.

07:50.795 --> 07:53.225
We saw this at the beginning
of the COVID crisis

07:53.225 --> 07:56.000
when certain breathing machines,

07:56.000 --> 07:58.490
the spare parts were not
easily available or they were

07:58.490 --> 08:01.310
far too expensive for
hospitals to purchase.

08:01.310 --> 08:05.060
A lot of 3D printers
or individuals

08:05.060 --> 08:06.710
using 3D printers created

08:06.710 --> 08:09.050
the spare parts to
sell to hospitals.

08:09.050 --> 08:11.870
However, it ended up in
several lawsuits because

08:11.870 --> 08:14.750
the original company
or the manufacturer of

08:14.750 --> 08:16.490
the breathing machines sued

08:16.490 --> 08:18.560
them for trademark infringement.

08:18.560 --> 08:22.280
You can also share plans
online with anyone

08:22.280 --> 08:25.640
else and allow them to print
whatever your designs are.

08:25.640 --> 08:27.080
This presents a challenge as

08:27.080 --> 08:29.935
even firearms can be
printed using 3D printing.

08:29.935 --> 08:31.700
However, I will tell
you this from most of

08:31.700 --> 08:32.940
the 3D printing of firearms

08:32.940 --> 08:34.850
if you were to look
online about this,

08:34.850 --> 08:36.620
most of these are
firearms you would

08:36.620 --> 08:38.450
not want to be holding
when you're shooting.

08:38.450 --> 08:40.760
There have been great strides

08:40.760 --> 08:41.840
in the improvement of these,

08:41.840 --> 08:44.330
but most of them
just can't handle.

08:44.330 --> 08:45.725
They're just not safe to be

08:45.725 --> 08:48.060
held while they're being shot.

08:48.880 --> 08:51.890
Let's summarize what we
went over in this lesson.

08:51.890 --> 08:54.125
We went over the
blockchain and its parts.

08:54.125 --> 08:57.050
We discussed artificial
intelligence and deep learning.

08:57.050 --> 08:58.880
We went over quantum computing,

08:58.880 --> 09:01.665
homomorphic encryption,
and 3D printing.

09:01.665 --> 09:04.205
Let's do some example questions.

09:04.205 --> 09:07.580
Question 1, what technology uses

09:07.580 --> 09:09.110
a peer-to-peer
network to distribute

09:09.110 --> 09:13.185
a ledger? The Blockchain.

09:13.185 --> 09:17.480
Question 2, what technology
deconstructs knowledge into

09:17.480 --> 09:19.280
a series of smaller parts so

09:19.280 --> 09:22.110
they can be used
to interpret data?

09:23.500 --> 09:28.025
Deep Learning. Question 3,

09:28.025 --> 09:32.730
what is the term for when AI
can be used against users?

09:34.450 --> 09:38.030
Adversarial artificial
intelligence.

09:38.030 --> 09:40.085
Finally Question 4,

09:40.085 --> 09:43.310
which technology allows for
the sharing of data for

09:43.310 --> 09:45.125
the purposes of analysis without

09:45.125 --> 09:48.030
actually revealing
the data itself?

09:49.130 --> 09:51.690
Homomorphic encryption.

09:51.690 --> 09:54.970
This brings us to
the end of Module 2.

09:54.970 --> 09:56.855
I hope you found
this module helpful,

09:56.855 --> 09:59.310
and I'll see you in Module 3.