There is much hype surrounding Artificial Intelligence (AI), but the most prominent consideration is that it is: “Artificial” Intelligence. That hype is a footprint consequence of a number of common concerns of the layman, outsider, enthusiast.
Some scrutiny undertakings:
- Will AI replace humans?
- What are the risks?
- AI lacks instinct, intuitiveness based on human nature vs. nurture tendencies and cultural habits and norms.
- AI can gain program knowledge for efficiencies, but can it gain and use knowledge intrinsically as do humans?
- “AI is capable of learning and adapting to new situations, but it lacks the emotional and intuitive capabilities of humans.”
A principle point for consideration is that AI is not only multifarious, but the capabilities of AI are stratospheric, e.g., small task/function bot operation to space voyaging and interplanetary spaceflight.
A Few “AI Key Terms Definitions” (Stanford University – Human-Centered Artificial Intelligence) (underlined examples added):
- Artificial Intelligence – a large class of software-based systems that receive signals from the environment and take actions that affect that environment by generating outputs such as content, predictions, recommendations, classifications, or decisions influencing the environments they interact with, among other outputs.
- Algorithm – a precise list of steps to take, such as a computer program. AI systems contain algorithms, but typically just for a few parts like a learning or reward calculation method.
- Deep Learning – the use of large multi-layer (artificial) neural networks that compute with continuous (real number) representations, similar to the hierarchically organized neurons in human brains. (see “neural network” narrative below)
- Foundation Models – an emerging class of models, often transformers trained by self-supervision on large-scale broad data, that can be easily adapted to perform a wide range of downstream tasks. The best-known examples are Large Language Models (LLMs), which focus on language-specific systems, but the term extends to models for all modalities of data and knowledge (such as images, videos, and audios). (e.g., natural language processing tasks, conversations: chats, GPT-3, Google Bard, etc.)
- Generative AI – a broad term to describe any AI system whose primary function is to generate content (including text, audio, codes, video, etc.).
- Machine Learning (ML) – a sub-field of AI that studies how computer systems can improve their perception, knowledge, decisions, or actions based on experience or data.
- Narrow AI – intelligent systems for particular tasks (e.g., speech or facial recognition). Human-level AI, or artificial general intelligence (AGI), seeks broadly intelligent, context-aware machines. It is needed for effective, adaptable social chatbots or human-robot interaction.
- Reinforcement Learning – a computer that enables autonomy by allowing an agent to learn action sequences that optimize its total rewards, such as winning games, without explicit examples of good techniques.
- Supervised Learning – a computer learns to predict human-given labels, such as particular dog breeds, based on labeled dog pictures.
- Unsupervised Learning – a computer that does not require labels to make predictions but sometimes adopts self-supervised learning, constructing its own prediction tasks such as trying to predict each successive word in a sentence.
Office of Management and Budget (OMB) Memorandum M-24-10, “Advancing Governance, Innovation, and Risk Management for Agency Use of Artificial Intelligence”, provides that, “The term “artificial intelligence” has the meaning provided in Section 238(g) of the John S. McCain National Defense Authorization Act for Fiscal Year2019,54 which states that “the term ‘artificial intelligence’ includes the following”:
- Any artificial system that performs tasks under varying and unpredictable circumstances without significant human oversight, or that can learn from experience and improve performance when exposed to data sets.”
All other definitions in the Memorandum begin with the term “an artificial system”, and then use the descriptive “requiring human-like perception”, “think or act like a human, including cognitive architectures and neural networks”, “machine learning”, “intelligent software”, “reasoning”, etc., references.
Pointedly, the Memo also includes this statement: “This definition of AI does not include robotic process automation or other systems whose behavior is defined only by human-defined rules or that learn solely by repeating an observed practice exactly as it was conducted.”
Hence, to first address some of the “scrutiny undertakings” above, it appears that AI can compute and perform autonomous outputs, demonstrate instinct and intuitive, knowledge-gaining, cognitive capabilities like-a-human. But, despite the sophistication, what is to be said about the fact that AI is “artificial”, and, the human “experience” in comparison, is unique?
But how does AI do this; how does the AI brain work?
The AI brain works through the use of “Neural Networks”.
Digging just a little deeper to provide a very basic explanation from a Carnegie Mellon University course: “TTOD 100 (G06N 3/02) Machine Learning Foundations – Deep Learning 1[i]”:
Neural networks function like the neurons of the brain. “The neural networks take in an input and generate an output. E.g., “Input: voice signal, Output: Transcription; Input: Image, Output: Text Caption; Input: Game state, Output: Next Move.”
Computing – In summary, (like the human brain, a mass of interconnected neurons), “the neurons connect in to other neurons and connect out to other neurons. The human brain is a connectionist machine. “Connectionism” is the key; the manner in which neurons are connected”. [ii]
An AI neural network performs simulated perception computations mimicking the way the brain works. The neural neuron receives inputs which also include varying “weights”, ( “a weighted sum of inputs”) to determine outputs of value or non-value to compute “perception”; this is a very basic view of how an AI brain works to compute results/make decisions and classifications; in other words, perform outputs. [iii] Complex inputs and connections can develop more sophisticated outputs.
AI also develops imaginary images; so, it appears that AI can “imagine” thru inputs and outputs.
“Neural Networks have become one of the main approaches to AI. They have been successfully applied to various pattern recognition, prediction and analysis problems. In many problems, they’ve established the state-of-the-art.”
AI is everywhere: In the cloud, PCs, Smart Phones, Gaming, Edge CPUs, etc.
What are the risks?
The fundamental risks are that AI must not be biased; it must be responsible, inclusive, accurate, transparent, and deployed to meet high standards.
THE FEDERAL ACQUISITION OF AI
The codified Federal Acquisition Regulation System of uniform policies and procedures, primarily through use of Federal Acquisition Regulation (FAR) prescribed contracting procedures Parts 8, 12, 13, 14 & 15 (and the relative provisions and clauses) is used to acquire and deliver on a timely basis the best value product or service for the government, while maintaining the public’s trust and fulfilling public policy objectives.
The same policies and contracting procedures have been and will be used to procure AI.
However, what additional laws, policies, etc., are established for AI?
Primarily,
- The AI in Government Act of 2020;
- The Advancing American AI Act;
- Executive Order 14110, Safe, Secure, and Trustworthy Development and Use of Artificial Intelligence.
Office of Management and Budget (OMB) Memorandums provide more specific instruction and implementation guidance to federal agencies:
- M-24-10 Advancing Governance, Innovation, and Risk Management for Agency Use of Artificial Intelligence,(this memorandum directs and provides requirements and guidance to federal agencies for the advancement, governance, and use of AI and AI innovation in the federal government while also managing risks, particularly those affecting the rights and safety of the public), and
- M-24-18 Advancing the Responsible Acquisition of Artificial Intelligence in Government,(this memorandum directs agencies to improve their capacity for the responsible acquisition of AI – provides for the institution of specific roles and responsibilities).
Specifically, as provided in OMB Memorandum M-24-18, when procuring AI in federal acquisitions, new guidance is provided primarily in three detailed considerations:
1 – Ensuring Cross-functional and Interagency Collaboration
Primarily the goal of this mandate is to engage agency and interagency enterprise-wide and cross-functional strategic and transactional planning and knowledge-sharing for the acquisition of AI to include the establishment of governance councils and working groups, revisions to existing policies and the development of new policies consistent with the existing laws, rules and regulations and directives.
2 – Manage Artificial Intelligence Risks and Performance
The use or adoption of AI poses risks to the protection of individual and entity rights, safety and information, if the AI is rights and safety-impacting. As stipulated in OMB Memorandum M-24-10, when risks result from reliance on AI outputs to make decisions, executive decisions and perform actions which could obstruct safety, equity, fairness, transparency and accountability, etc., (e.g., AI used for more autonomous outputs), agencies must eliminate and/or mitigate those risks, and through the contracting process. Moreover, when contracting for AI, agencies must ensure that civil rights and liberties are protected and that AI Biometrics (e.g., biometric identifiers: faces, irises, fingerprints, etc.) protect the public’s rights, safety and privacy. Cybersecurity considerations must also be employed when contracting for AI. Protective measures must be employed and AI must be responsible.
“Rights-Impacting and Safety-Impacting AI: AI whose output serves as a basis for decision or action that has a legal, material, or similarly significant effect” on a variety of rights and privileges.”
3 – Promoting a Competitive AI Market with Innovative Acquisition
The overarching emphasis is to employ best practices and innovative best practices and competitive procedures to promote AI interoperability, prevent vendor lock-in and avoid limiting future options while consistently procuring state-of-the-art AI.
Will AI replace humans?
The Prevalence of AI
In fact, AI has been used/deployed for decades; NASA is an outstanding example. Does the augmentation of information availability require increased economical efficiencies? There is an increased societal focus on improving productivity, automating processes, and modernizing experiences.
How practical is AI? The answer lies in “Use Cases” – simply stated, they are real world cases/situations/examples of how AI has been used; AI deployed developments. Specifically, for example, Generative AI is “used” in all sectors of society: healthcare, finance, transportation, retail, government, etc.
“AI Agents” (Generative AI) are used in these sectors as Customer agents (salesperson output function and purpose); Employee agents (administrative/personal/virtual assistant output function and purpose); Data agents (data and research analyst output function and purpose); Creative Agents (design and production output function and purpose), etc.
However, AI deployment in any sector is dependent upon specific needs, efficiencies, risks, presumptive risks, affirmatives for risk-mitigation, past performance, interoperability, cost efficiency, etc.
Some comments concerning AI:
A neurosurgeon from MIT stated that human brain neurons will never homogenize or fuse with the brain neurons of AI
“As robots working alongside humans become smarter and smarter, humans working with them will naturally think of them as co-workers. As these AI-enable robots become more and more autonomous, they may develop a desire to be treated the same way as their human coworkers.”
“The central problem with grounding AI rights in empathy, then, is that empathy can easily push one’s moral decisions in the direction of injustice. It is not only harder for us to empathize with people who are different from us, but it is easy to make mistakes when we attempt to put ourselves in the shoes of another.”
“The ‘Hard Problem’ for AI rights, I contend, stems from the fact that we still lack a solution to the ‘Hard Problem’ of consciousness”….. “why certain functions or brain states are ‘accompanied by experience’ ”
AI’s misuse can infringe on human rights by facilitating arbitrary surveillance, enabling censorship and control of the information realm, or by entrenching bias and discrimination.”
Summary consideration:
AI requires programmable human experience and knowledge for simulation/output. There are subtle, human idiosyncrasies, natural and nurtured evolving environmental and cultural tendencies and their corelating and related behaviors that AI cognitive design cannot timely or appropriately identify or assimilate for true accuracy. AI feelings are not human feelings; AI logic is not human logic that is developed through unique human experience.
Can AI simulate “Appreciation” and “enjoyment” at all, or, of the human feeling experience?
To what extent and what are the pathways and expressions of appreciation?
What are the impacts of the human experience on the brain when exercising, eating nutritious food, etc., and how would these AI programmed or learned experiences or outputs be “real”?
Consider this –
In a lecture concerning the Holy Writings and Creator of the Universe, reference was made to scientists and astronomer’s observations and estimations and facts and information contained in the encyclopedia (you can also google this info):
- The universe is constantly expanding.
- The Milky Way Galaxy alone contains some 200 to 400 billion stars.
- The solar system moves at a speed of 220 kilometers per second, and 568,000 miles per hour.
- A light beam takes 100,000 years to travel from one end of our Milky Way Galaxy to the other.
- There are an estimated 100 billion to 1 trillion Galaxies in the universe.
- The estimated number of stars in all of the Galaxies: 10 Sextillion to 1 Septillion
What is the universe of the human brain?
- Weight of human brain 3lbs;
- Contains nearly100 billion neurons; 100 trillion synaptic connections
- Most complicated organ in the universe
- It is said that there are as many neurons in the human brain as there are stars in the Milky Way Galaxy.
Will Artificial Intelligence replace humans?
Apparently, the human brain is used to create “Artificial Intelligence”
What would you say?
_______________________________________________________________________
[i] Professor Bhiksha Raj
[ii] Alexander Bain – “Mind and Body” 1873, “The information is in the connection”.
[iii] Mimics “Boolean” gates, expressions and functions.
The Public Contracting Institute (PCI) offers training on Artificial Intelligence.
Join Tyler Evans, of Steptoe, on December 5, 2024 for AI in National Security & Defense Procurement. If you don’t find that interesting, consider joining Jim Goepel, of the CMMC Information Institute, for Generative AI: You Know Better Than to Trust A Strange Robot on May 13, 2025.
