SCINDIA RESEARCH AND/OR STUDY GUIDE

Artificial Intelligence (AI) Artificial Intelligence (AI) signifies a significant shift in technology, enabling
machines to perform tasks that traditionally required human intelligence. This section delves into the
various aspects of AI, shedding light on its complexities and controversies.

Types of AI

AI can be divided into two main types: Narrow AI and General AI. Narrow AI, also known as Weak AI, is
designed to perform specific tasks or solve specific problems, such as speech recognition or image
classification. General AI, on the other hand, refers to AI systems with human-like cognitive abilities,
capable of understanding and reasoning across a wide range of tasks

Key Components of AI

AI encompasses several core technologies that enable machines to exhibit intelligent behavior:

● Machine Learning: A subset of AI that enables systems to learn from data and improve their
performance over time without explicit programming.

● Natural Language Processing (NLP): The ability of machines to understand and generate human
language, enabling applications such as virtual assistants and language translation.

● Computer Vision: AI algorithms that enable computers to interpret and analyze visual information
from images or videos, facilitating applications like object detection and facial recognition.

● Robotics: The integration of AI and robotics enables machines to interact with the physical world,
performing tasks such as manufacturing, healthcare assistance, and autonomous navigation

Applications of AI AI technologies have found applications across various sectors, revolutionizing

industries and transforming everyday experiences:

● Personalized Experiences: AI personalizes experiences across various sectors. From recommendation

engines in e-commerce suggesting products you might like to AI-powered news feeds filtering content
based on your interests, AI tailors experiences to individual preferences.

● Smart Automation: AI automates tasks in numerous fields. For instance, AI-powered chatbots handle
customer service inquiries, and AI algorithms manage complex logistics networks in the supply chain.

● Medical Diagnosis and Treatment: AI assists healthcare professionals in medical diagnosis through
image analysis and pattern recognition in patient data. It also aids in drug discovery and development of
personalized treatment plans.

● Financial Services and Fraud Detection: AI helps financial institutions assess risk, prevent fraud, and
personalize financial products. AI algorithms can analyze vast amounts of financial data to detect
suspicious activity and prevent financial crimes.

● Self-Driving Cars and Advanced Driver-Assistance Systems (ADAS): AI is at the forefront of developing
self-driving cars and ADAS systems that enhance safety and automate driving functions.
 ● Cybersecurity Threat Detection and Prevention: AI plays a crucial role in cybersecurity by analyzing
network traffic and identifying potential threats in real-time. AI-powered systems can detect and
respond to cyberattacks more effectively.

● Content Creation and Marketing: AI assists in content creation by generating ideas, writing different
creative formats of content, and optimizing content for search engines. AI also helps target marketing
campaigns to specific audiences based on their online behavior.

● Manufacturing and Quality Control: AI is used in predictive maintenance to prevent equipment

failures and improve overall production efficiency. AI-powered systems can also automate quality control
processes in manufacturing.

● Scientific Research and Drug Discovery: AI accelerates scientific research by analyzing vast datasets and
identifying patterns that might lead to breakthroughs. AI aids in drug discovery by simulating molecular
interactions and designing new drug candidates.

● Climate Change and Environmental Monitoring: AI is used to analyze climate data, predict weather
patterns, and monitor environmental changes. AI models can help us understand the impact of climate
change and develop sustainable solutions

Ethical Considerations

As AI technologies become increasingly integrated into society, ethical considerations and societal
impacts become paramount. Issues such as algorithmic bias, privacy concerns, and the ethical use of AI
in decision-making processes must be carefully managed to ensure fairness, transparency, and
accountability. By gaining a deeper understanding of the multifaceted nature of AI, stakeholders will be
better equipped to engage in meaningful conversations and propose effective regulatory measures to
govern its development and deployment.

An AI system is a machine-based system that, for explicit or implicit objectives, infers,

from the input it receives, how to generate outputs such as
predictions, content, recommendations, or decisions that can influence physical or
virtual environments. Different AI systems vary in their levels of autonomy and
adaptiveness after deployment
Topics typically encompassed by the term “AI” and in the definition of an AI system
include categories of techniques such as machine learning and knowledge-based
approaches; application areas such as computer vision, natural language processing,
speech recognition, intelligent decision support systems, and intelligent robotic systems;
and specific applications of these tools in different domains.
The OECD just released an explanatory memorandum on its updated definition of an AI
system. In this blog post, we explain the main points.

Input, including data

Input is used both during development and after deployment. Input can take the form of
knowledge, rules, and code that humans put into the system during development or
data. Humans and machines can provide input. During development, input is leveraged
to build AI systems, e.g., with machine learning that produces a model from training
data and/or human input. Input is also used by a system in operation, for instance, to
infer how to generate outputs. Input can include data relevant to the task to be
performed or take the form of, for example, a user prompt or a search query.

Illustrative, simplified overview of an AI system

Note: This figure presents only one possible relationship between the development and
deployment phases. In many cases, the design and training of the system may continue in
downstream uses. For example, deployers of AI systems may fine-tune or continuously train
models during operation, which can significantly impact the system’s performance and
behaviour.

Types of AI systems and how they are built

The definition that OECD countries agreed to use describes characteristics of machines
considered to be AI, including what they do / how they are used but also how they are
built.
Although different interpretations of the word “model” exist, in this document, an AI
model is a core component of an AI system used to make inferences from inputs to
produce outputs. Prior to deployment, an AI system is typically built by combining one or
more “models” developed manually or automatically (e.g., with reasoning and decision-
making algorithms) based on machine and/or human inputs/data.
Explanatory memorandum on the updated OECD definition of an AI system
 Machine learning is a set of techniques that allows machines to improve their
performance and usually generate models in an automated manner through
exposure to training data, which can help identify patterns and regularities, rather
than through explicit instructions from a human. The process of improving a
system’s performance using machine learning techniques is known as “training”.

 Symbolic or knowledge-based AI systems typically use logic-based and/or

probabilistic representations, which may be human-generated or machine-
 generated. These representations rely on explicit descriptions of variables and of
their interrelations. For example, a system that reasons about manufacturing
processes might have variables representing factories, goods, workers, vehicles,
machines, and so on.

 In addition, it should be noted that symbolic AI may use machine learning. For
example, inductive logic programming learns symbolic logical representations
from data, and decision-tree learning learns symbolic rules in the form of a tree of
logical conditions.
People tend to consider that AI systems built using machine learning have some
intelligence taking place because, during their training, they figure out relationships
between model parameters without precise instructions. For example, language models
are given large amounts of language resources and given the objective to “predict the
next word”, producing a trained model that appears to respond intelligently to prompts .
So, this is where what looks to AI researchers as intelligence occurs: in the building
phase. What looks to end users like “intelligence” occurs at runtime, after construction.
But when building a symbolic AI system manually, humans supply the knowledge and
the vocabulary in which it is expressed. Here, we consider the knowledge engineer to
be the source of part of the intelligence – i.e., the AI system does not discover the
knowledge it uses from its own experience. On the other hand, the AI system may
perform very complex reasoning that contributes to the overall intelligence of the
system. For example, Deep Blue, like AlphaZero, is given the rules of the game by
human engineers; those rules allow it to play legal moves, but not good moves. The
good moves come largely from its prodigious ability to reason about future trajectories in
the game.

RELATED >> Updates to the OECD’s definition of an AI system explained

Autonomy and adaptiveness

Explanatory memorandum on the updated OECD definition of an AI system
An AI system’s objective setting and development can always be traced back to a
human who originates the AI system development process, even when the objectives
are implicit. However, some AI systems can “adapt” or develop implicit sub-objectives
and sometimes set objectives for other systems. Human agency, autonomy, and
oversight vis-à-vis AI systems are critical values in the OECD AI Principles that depend
on the context of AI use.
 AI system autonomy means the degree to which a system can learn or act without
human involvement following the delegation of autonomy and process
automation by humans. Human supervision can occur at any stage of an AI
system’s lifecycle, such as during AI system design, data collection and
processing, development, verification, validation, deployment, or operation and
 monitoring. Some AI systems can generate outputs without specific instructions
from a human.
 Adaptiveness, contained in the revised definition of an AI system, is usually
related to AI systems based on machine learning that can continue to evolve
their models after initial development. Examples include a speech recognition
system that adapts to an individual’s voice or a personalised music recommender
system. AI systems can be trained once, periodically, or continually. Through
such training, some AI systems may develop the ability to perform new forms of
inference not initially envisioned by their developers. The concept of post-
deployment adaptation is highly significant for the regulation of AI because it
means that assurances concerning the performance and safety of the system at
the time of deployment may be invalidated by subsequent adaptation. Thus,
assurances must be obtained for all future versions of the system under all
possible future data trajectories, which is a considerably more difficult problem
than assuring a static system. This difficulty may be alleviated to some extent by
incorporating automated testing as part of the adaptation process so that, for
example, no bias creeps into a system that is initially certified as fair.

The OECD definition of an AI system intentionally does not address the issue of liability
and responsibility for AI systems and their potentially harmful effects, which ultimately
rests with humans and does not in any way pre-determine or pre-empt regulatory
choices made by individual jurisdictions in that regard.
The updated definition of AI is inclusive and encompasses systems ranging from simple
to complex. The fact that a system is “simple” does not mean that it carries no risk or
that its safety need not be assured, but it does mean that providing such assurances
may be simpler than it would be for a complex system! AI represents a set of
technologies and techniques applicable to many different situations. Specific sets of
techniques, such as machine learning, may raise particular considerations for
policymakers, such as bias, transparency, and explainability, and some contexts of use
(e.g., decisions about public benefits) may raise more significant concerns than others.
Therefore, when applied in practice, additional criteria may be needed to narrow or
otherwise tailor the definition when used in a specific context, and additional regulations
may apply to certain types of AI systems, even in the same context of use.

AI system objectives
Explanatory memorandum on the updated OECD definition of an AI system
AI system objectives can be explicit or implicit. For example, they can belong to the
following categories that may overlap in some systems:
 Explicit and human-defined. In these cases, the developer encodes the objective
directly into the system, e.g., through an objective function. Examples of systems
with explicit objectives include simple classifiers, game-playing systems,
reinforcement learning systems, combinatorial problem-solving systems,
 planning algorithms, and dynamic programming algorithms.

 Implicit in rules and policies. Rules, typically human-specified, dictate the action
to be taken by the AI system according to the current circumstance. For example,
a driving system might have a rule, “If the traffic light is red, stop.” However,
these systems’ underlying objectives, such as compliance with the law or
avoiding accidents, are not explicit in the system, even if they are apparent to
and intended by the human designer.

 Implicit in training data. This is where the ultimate objective is not explicitly
programmed but incorporated through training data and a system architecture
that learns to emulate those data, e.g., training large language models to imitate
human linguistic behavior. In such cases, the human engineer may not know
what objectives are implicit in the data.

 Not fully known in advance. Some systems may learn to help humans by learning
more about their objectives through interactions. Some examples include
recommender systems that use “reinforcement learning from human feedback” to
gradually narrow down a model of individual users’ preferences.

 Bias in Algorithmic Decision-Making: AI algorithms are trained on massive

datasets, and if those datasets contain biases, the algorithms can perpetuate
those biases. For example, an AI system used for loan approvals might deny
loans to people from certain neighborhoods or with certain names.
 Privacy Violations and Data Exploitation: Big data companies collect vast
amounts of personal information about users. This data can be used for targeted
advertising, but it can also be misused for identity theft, discrimination, or even
physical harm.
 Lack of Transparency and Explainability: Many AI systems are complex
"black boxes" where it's difficult to understand how they make decisions. This
lack of transparency can make it hard to identify and address bias or errors in the
system.
 Algorithmic Manipulation and Social Media Addiction: Social media
platforms can use AI to personalize user feeds and content, potentially creating
echo chambers and promoting addictive behaviors. This can lead to the spread
of misinformation and the manipulation of public opinion.
 Job displacement without proper reskilling: AI automation is already
replacing some jobs, and this trend is likely to continue. Big data companies have
a responsibility to help prepare the workforce for these changes, but this is not
always happening.

Here are some specific examples of big data companies being accused of AI misuse:

 Amazon's facial recognition technology Rekognition: This technology has

been criticized for showing racial bias in identifying people.
 Facebook's newsfeed algorithm: Facebook has been accused of using its
algorithm to promote fake news and filter bubbles, contributing to political
polarization.
 Cambridge Analytica scandal: This company harvested data from millions of
Facebook profiles without user consent and used it to target voters in political
campaigns.
Penspark

EHMUN BG
 Moreover, with the advent of AI technologies, the creation of deep fakes has
become a
relatively easy task, complicating this issue of misinformation further. For
instance, an
incident involving the use of a fake voice of President Joe Biden in robocalls
aimed at
deterring Democrats from participating in a primary. This indeed showcases the
incredible
potential for deepfakes to interfere in elections. Similarly, AI images that falsely
depicted
former President Donald Trump sitting with teenage girls on Jeffrey Epstein’s
plane
circulated on social media. A deep fake posted on X last February portrayed a
leading
Democratic candidate for mayor of Chicago as indifferent toward police
shootings.
Privacy breaches represent another issue here, with incidents often exposing vulnerabilities

in how social media platforms manage user data. For example, allegations of Meta's

interference in elections, notably raised by Donald Trump, showed broader concerns about

the impact of social media on politics and election integrity; these claims, however, were

later debunked and led to legal consequences for Trump and his allies. Meta, in particular

though, has faced intense scrutiny and legal challenges over data privacy issues, which

were exemplified by the Cambridge Analytica scandal. Under this, Meta was accused of

misleading shareholders and users about the risks to user data. This led to a U.S. appeals

court reviving a class action lawsuit against the company and resulted in Meta paying over

$5 billion in penalties to U.S. authorities, in addition to a $725 million settlement to users.

Deepfakes

These refer to electronically and artificially manipulated images or videos made using artificial

intelligence technologies that convincingly replace a person’s face or voice in an existing video

with someone else’s. Deepfakes pose a few threats to people that the Senate needs to

address.

Political Disinformation: Deepfakes enable people to manipulate what famous political

figures say and even who is saying it. This is harmful because news is propagated on social

media easily, and people tend to believe what they see if it’s posted by someone they know,

especially since deep fakes are convincing. BuzzFeed posted a video of President Obama
saying something he never did, and apps like FakeApp make it easier for every person to do the

same thing. This causes geopolitical and domestic tensions, which cause unrest between

nations, between people, or between the state and the people. This is also problematic

because it decreases trust in the government, leading to political unrest and people not

believing in state legislation, bodies, and the judiciary, which are all harms we don’t want. The

state loses credibility, and cynicism within the system increases, which is harmful.

Defamation: Increasingly, fake videos of celebrities and other famous personalities

face problems with people using malicious deep fake software to put them in videos

doing things they never did or audio saying things they never did. This is bad because it

ruins their public image and personal lives, and people tend to believe things about

them that are malicious. Tom Cruise was impersonated after a fan created an account

with deepfakes of his, which quickly went viral on the internet. Even though this was to

prove AI’s capabilities, the point is that it’s increasingly accessible and effective. Note

that it’s increasingly difficult to litigate cases against deepfakes, considering the

anonymity tied to platforms like Instagram or TikTok, which don’t require geographical

location or proximity to post on them.

Cyberwarfare

This refers to the use of digital attacks, spread specifically on social media, by a country or

organization to disrupt the vital systems of another or cause unrest, intending to create

damage. There are a few risks associated with cyber warfare.

Weaponization of social media: Social media platforms are steadily becoming persuasive

tools of propaganda. Institutions like ISIS and other players use false news or bogus

accounts to stoke fear, incite violence, or manipulate outcomes. This is easily persuasive

and harmful because it oversaturates the news on social media that is useful and spreads

misinformation to vulnerable people.

Social media recruitment: In 2014, ISIS changed its strategy from being confidential to

spreading propaganda and their motives on Twitter (now X). This perpetuates narratives

about their work and manipulates people into supporting acts of terrorism. This is especially

harmful when it reaches the realm of national unrest when people are swayed to show
 dissatisfaction in circumstances of war. They can use social media to create a sense of

community and belonging for potential recruits and to glorify violence and extremism.

The Emerald Heights Model United Nations 2024 US SENATE | 14

During the 2016 U.S. presidential election, Russia launched a social media campaign to spread

misinformation about the candidates and undermine the democratic process.

A Russia-linked troll farm called the Internet Research Agency (IRA) was believed to be funded

by a close associate of Vladimir Putin. The IRA created a set of fake accounts that seemed

real, posing as American citizens from different locations in the USA with different interests.

These accounts were used by the IRA to flood social media platforms and artificially sow

distrust in the people; this included pro-Trump propaganda, anti-Clinton propaganda, and

highly divisive content on social issues like gun control and immigration. All of this was further

polarized using specialized algorithms that targeted specific users and demographics, focusing

on users’ political interests and geographical locations. This content was designed to elicit

emotions of fear, anger, and unrest in the people, and it worked. This highly emotional content

was more likely to be shared and further amplified, reaching a wider audience. This created a

climate of suspicion and exacerbated existing political divides.

NVIDIA has made several commitments and taken steps towards openness in various
aspects of AI development. Here's a breakdown of some key areas:

Open Source Hardware and Software:

 CUDA Toolkit and Libraries: NVIDIA provides open-source tools and libraries
like CUDA that allow developers to leverage the power of NVIDIA GPUs for
general-purpose computing and AI applications.
 TensorRT: This is an open-source inference optimizer that helps developers
deploy trained AI models for efficient performance on NVIDIA hardware.
  Open-Source Hardware Designs: While NVIDIA doesn't entirely open-source
their hardware designs, they do participate in collaborative projects and
contribute to open hardware standards.

Research and Development Transparency:

 Open Research Publications: NVIDIA researchers frequently publish papers

detailing their latest advancements in AI algorithms, hardware architectures, and
software tools. This contributes to the overall knowledge base in the field.
 Participation in Open Science Initiatives: NVIDIA scientists participate in
workshops and conferences, sharing their research findings and fostering
collaboration with the broader scientific community.
 Open Datasets: NVIDIA has released some open datasets for specific research
areas like self-driving cars and natural language processing, which can benefit
researchers worldwide.

Educational Resources and Developer Support:

 NVIDIA Developer: This is a comprehensive platform offering tutorials,

documentation, code samples, and other resources to help developers learn and
build AI applications using NVIDIA technologies.
 Deep Learning Institute (DLI): NVIDIA offers free online courses and
workshops on deep learning concepts and programming with their tools.
 Teaching Materials and Curriculums: They collaborate with universities and
educational institutions to provide AI learning materials and support the
development of AI curriculums.

It's important to note that Openness in AI is a complex issue. While NVIDIA has
made strides in these areas, there's still room for improvement. Some ongoing
discussions include:

 Level of Hardware Openness: While some aspects are open-source, the full
schematics of NVIDIA GPUs remain proprietary.
 Accessibility of Resources: Balancing open access with the need to protect
intellectual property can be challenging.

Overall, NVIDIA's initiatives demonstrate a commitment to openness in AI development.

They contribute to advancements in the field by providing tools, fostering collaboration,
and promoting education.

 Bias in Algorithmic Decision-Making: AI algorithms are trained on massive

datasets, and if those datasets contain biases, the algorithms can perpetuate
those biases. For example, an AI system used for loan approvals might deny
loans to people from certain neighborhoods or with certain names.
 Privacy Violations and Data Exploitation: Big data companies collect vast
amounts of personal information about users. This data can be used for targeted
 advertising, but it can also be misused for identity theft, discrimination, or even
physical harm.
 Lack of Transparency and Explainability: Many AI systems are complex
"black boxes" where it's difficult to understand how they make decisions. This
lack of transparency can make it hard to identify and address bias or errors in the
system.
 Algorithmic Manipulation and Social Media Addiction: Social media
platforms can use AI to personalize user feeds and content, potentially creating
echo chambers and promoting addictive behaviors. This can lead to the spread
of misinformation and the manipulation of public opinion.
 Job displacement without proper reskilling: AI automation is already
replacing some jobs, and this trend is likely to continue. Big data companies have
a responsibility to help prepare the workforce for these changes, but this is not
always happening.

Here are some specific examples of big data companies being accused of AI misuse:

 Amazon's facial recognition technology Rekognition: This technology has

been criticized for showing racial bias in identifying people.
 Facebook's newsfeed algorithm: Facebook has been accused of using its
algorithm to promote fake news and filter bubbles, contributing to political
polarization.
 Cambridge Analytica scandal: This company harvested data from millions of
Facebook profiles without user consent and used it to target voters in political
campaigns.

or the Regulation of AI:

 Focus on Specific Areas:

o Regulating AI in high-risk applications (e.g., autonomous vehicles, medical
diagnosis)
o Addressing bias and fairness in AI algorithms
o Establishing transparency and accountability for AI decision-making
 Standards and Oversight:
o Developing national or international standards for AI development and
deployment
o Creating oversight bodies to monitor AI use and enforce regulations
o The role of government vs. private sector in AI regulation
 Impact on Jobs and the Workforce:
o The potential for job displacement due to AI automation
o Preparing the workforce for jobs in the AI economy
o Ethical considerations of AI replacing human jobs

For the Role of Big Data Companies:

 Data Privacy and Security:

 o Strengthening data privacy laws and user control over personal
information
o Ensuring the security of big data from breaches and misuse
o The role of big data companies in protecting user privacy
 Competition and Antitrust Concerns:
o Addressing the dominance of big data companies and potential for anti-
competitive practices
o Ensuring fair access to data for smaller companies and innovation
o The role of government in regulating big data monopolies
 Social Responsibility and Algorithmic Bias:
o Holding big data companies accountable for the societal impacts of their
algorithms
o Mitigating bias in data collection and algorithmic decision-making
o The responsibility of big data companies to promote ethical AI
development

Additional Considerations:

 International Cooperation: The need for international collaboration on AI

regulation and big data governance.
 The Future of AI: Discussing long-term considerations for AI development and
potential risks.
 Public Education and Awareness: Raising public awareness of AI and big data
issues, and fostering informed discussions on responsible development.

By using these topics as starting points, you can create a productive and e