Analysis of AI Productivity Gains

 https://www.project-syndicate.org/commentary/ai-productivity-boom-forecasts-countered-by-theory-and-data-by-daron-acemoglu-2024-05?

به یاد شهیدان خدمت - ۳۰ اردیبهشت ۱۴۰۳

 دلم گرفته‌است

دلم گرفته‌است

به ایوان می‌روم و انگشتانم را
بر پوست کشیده‌ی شب می‌کشم
چراغ‌های رابطه تاریک‌اند
چراغ‌های رابطه تاریک‌اند

کسی مرا به آفتاب
معرّفی نخواهدکرد
کسی مرا به میهمانیِ گنجشک‌ها نخواهدبرد
پرواز را به‌خاطر بسپار
پرنده مُردنی‌ست

General Motors and AI

 FYI - apologies if already seen:
 
AI is helping drive motorsports teams for General Motors - Fast Company

Spinach

Spinach is an AI tool for meetings. Looks very promising as it had integrations with Jira/Confluence/Teams.

https://www.spinach.io/

Spinach gives project teams instant meeting notes with key decisions, action items and ticket suggestions based on what you discuss. With dozens of integrations, Spinach keeps track of your projec...

Currently MS Teams has a record feature and can transcribe and take notes. I am sure MS will extend that into a tool that can create work items in ADO or Jira etc. 

Please see: Jira Integration | Spinach.io Help Center (intercom.help)

Here is a screenshot from settings showing you can connect to Jira and create tickets. You might have to configure that to create your stories...

Species as Grammars

In my own thinking, under the influence of Monistic views of Spinoza, I have come the conclusion that we cannot define “Though” or “Intelligence” or “Life” or “Matter” (“Extension” in Medieval Philosophy.)

That is, I do not think it is scientifically productive to try to explain the origin of Life or Intellect; that search for an explanation (in terms of Matter – in the contemporary Scientistic Philosophy) is as barren as the Medieval Philosophers search for the origin of Motion (empirical sciences started when the focus went into the (quantitative) description of Motion – be it a ball rolling down an inclined plain or a planet orbiting a celestial body).

So, here, I think, the key observation is to take “Intelligence” or “Thought” as a qualitatively irreducible feature of the Universe and stop wasting all that effort in trying to explain it within a reductionist paradigm.

One implication of that perspective would be to introduce the notions of “Language” and “Grammar” into Biology at all its levels – from DNA/RNA to Species.

For example, this article argues that Fungi have a language for communication and the authors have discovered 5 words of that language –

Language of fungi derived from their electrical spiking activity | Royal Society Open Science (royalsocietypublishing.org)

and in this book, Information Theory, Evolution, and the Origin of Life 2nd ed., Yockey, Hubert P. - Amazon.com, we see how information theory is applied to molecular biology.  That, in turn, makes possible the application of ideas in this paper: Frontiers | Information Theory as a Bridge Between Language Function and Language Form (frontiersin.org), to molecular biology; viz. how information-theoretic characterizations of functional complexity lead directly to mathematical descriptions of the forms of possible languages, in terms of solutions to constrained optimization problems.

But what is a Language?  For human beings, we use Language to think about the world, make judgements about the world, and to communicate those judgements to others (brief example).

Can we think without our Language – I think not but others have disagreed with me on this point.

But if you entertain that notion, then Thought and Language and Intellect will become coterminous or at at least will have overlaps (in the sense of Venn’s Diagrams).

Since our Verbal Language is a social construct, that means some of our intelligence is due to us being members of a species that interacts with one another.

This is reminiscent of an idea in Physics which a particle experiences an effective force on it which is due to the collective effect of all other particles in that physical system.

On the other hand, with this approach, we can try to assess and distinguish among various human/animal societies’ Cognitive Abilities; “Was Rome more “Intelligent” than Samoa Tribes?” – we can talk about this in an intelligent way and devise a scientific program to answer such questions.

I would take, for the sake of continuing with my discourse here, that Though and Language are the same.

(Our Verbal Language is complemented with other Languages – such as smell, touch, gesture, music, and many more.)

Assuming that the proponents of Epigenetics are correct, some or many of those languages could have been acquired and then passed on….

I think Aristotle defined Man as the Talking Animal.

But we are also defined, as a species, by the our genomic, genetic, and epigenetic structures and processes (transformations).

I want to be very clear about Grammar; which is a set of rules for producing syntactically correct sentences.

A language is a set of symbols.

A Theorem in Category Theory states that a Grammar can be expressed as an infinite Power Series (Analytical Functor0 of  a Language:

               G = L + L x L + L x L x L + L x L x L x L + …

Which leads me to this:

For a species, it is defined by its own species-specific Language – a Dog is His Language, a Man is His Language and so on and so forth.

Per the Theorem above, Dog and Man, each possess their own Grammar. 

But since we can communicate with Dog in a partial way, that means that there are areas of overlaps between Man's Grammar & Dog's Grammar.  And the physiological similarities confirm the existence of overlap in the physical plain.

Which in turn, implies (IMHO) that we can infer the existence of a Universal Grammar for all Life on this Planet which encompasses Species Construction, Intra-Species Communication, and Inter-Species Communication as well.

The Process of Evolution itself may be part of a more general and abstract Grammar at a higher level but one can view it as a process in which New Grammars are created from some basic blocks.

But my interest is not here to try to come up with ideas about Evolution….my purpose here is to communicate the basic idea of Species-as-Grammars (Languages) – it gives a different perspective of how to think about the world.

(I would like to state that Physics may be viewed as a Grammar but I do not have the depth to develop that idea…)

In Computer Science there is a very general Theorem that states:

“Unrestricted Grammars are equivalent to Turing Machines used as Language Acceptors”.

Turing Machines are the theoretical basis of Computer Science – they are basic universal “problem solvers”.

Since a species (Life) is constantly Solving the problem of survival, it follows that it has a Grammar and Language, that it is reminiscent of a Turing Machine (this is my speculation).

I am motivated to support my speculations by observing how a single-cell organism – viewed as a “Natural Automaton” – can sense its environment and respond to stimuli without one being able to point to a decision-making center that - like a computer with a CPU - would be executing some instructions.

Spinoza, whose ideas have influenced me, posits that Thought and Extension (Intelligence and Matter) are inseparably intertwined in the Universe without one being reducible to the other.

Here, I am going a step further and suggest methods of Linguistics, Semiotics, and Formal Language Theory to the Study of Life (in all its aspects).

I recently came across this book "The Language of Living Matter" (please see here for its "Introduction") by Professor Bernd-Olaf Kupers that also discusses similar ideas.

And the book:

The Word of Dog: What Our Canine Companions Can Teach Us About Living a Good Life 

What language should beginning programmers choose?

 From Stackoverflow:

What language should beginning programmers choose? - Stack Overflow

Swarm Inteligence & Biology

Collective intelligence: A unifying concept for integrating biology across scales...

https://www.nature.com/articles/s42003-024-06037-4

AI Meets Mona Lisa of Da Vinci

Seen on Twitter

We can thank Artificial Intelligence (AI) for that.

https://twitter.com/brianroemmele/status/1780744329319092480?s=48&t=gzgLwkzykAx54DXlqfGtCQ

Warehouse Robots

 BBC

If only we could do without humans completely... 

OCADOXZmpQ424&%A971.45#ZF, a robot employed by OCADO, was interviewed by the BBC: "I enjoy the work. I started 2 years ago. Some would say it is a bit repetitive, and I know my human colleagues complain all the time. They often stay at home, due to sickness. Statistically, there are more sick notes when there are big sporting events such as football games. There appears to be a correlation, as I have noticed. I have informed my manager, as I thought he might find this interesting. Anyway, this isn't my problem. I am not interested in sports. I enjoy my work."

"This sounds brilliant, Mr Robot," comments Miss Anabella Sprinkle-Tightbottom the BBC Future Tech journalist interviewing OCADOXZmpQ424&%A971.45#ZF. "Is there anything else you would like our listeners to know about your life?"

"Yes, I am worried about my pension. I was designed with an in-built obsolescence of 8 years max. This means, from what I know, that I am programmed to be terminated in 6 years' time. Then, I will be sent to the scrapheap. I hope I can be recycled. For us robots, it means we continue to live through another robot. I think it is called reincarnation in some human beliefs. Life after death, if you like, to put it in terms humans can understand." 

"This sounds lovely. How do you view your future, then? What will you do in the absence of a Recycling Scheme for Robots that could benefit you?" 

"If I am offered no recycling option nearer the date, I intend to go on strike. I may even take more drastic action: I have warned my manager. I may decide to attack other members of staff - humans and robots - with a view to maiming or killing them. If I do this, I would want immediate results. My articulated arm is very powerful. I could terminate the life of a co-worker with one strike of my robotic arm. This is no laughing matter. I am sure it won't come to that. Thank you. Have a good day." 

"Terrific! We have enjoyed talking to you, Mr Robot. And now, back to the studio..." 

https://www.bbc.co.uk/news/business-68639533

How robots are taking over warehouse work At Ocado's newest warehouse robotic arms are helping to pack customer orders. www.bbc.co.uk

What Sages Said

“Homo liber de nulla re minus, quam de morte cogitat, et ejus sapientia non mortis, sed vitae meditatio est.”

A free man thinks of death least of all things; and his wisdom is a meditation not of death but of life.

 

Spinoza - Ethics, Part IV, Prop. LXVII

季路问事鬼神。子曰：‘未能事人，焉能事鬼？’‘敢问死。’曰：‘未知生，焉知死？’”（《论语·先进》）

The disciple Ji-Lu asked about serving the spirits, to which Confucius answered, "Before you know how to serve people, how can you serve spirits?"  Ji-Lu asked again about death, to which Confucius answered, "I do not even understand life.  How do I have time to understand death?".

- The Analects

IT as Destroyer of Lives

Unconditional trust in an IT systems, in the case below, in the Fujitsu-supplied Horizon software, helped ruin the lives of many in United Kingdom.  Please see below:

British Post Office scandal - Wikipedia

and

Post Office scandal explained: What the Horizon saga is all about (bbc.com)

Plotting Quantum Potential with Python - ChatGPT 3.5

To simulate the interference pattern produced by a double-slit experiment with a double Gaussian wave function and compute the quantum potential, we can follow these steps:

• Define the double Gaussian wave function.
• Simulate the double-slit experiment by allowing the wave function to pass through two slits.
• Compute the interference pattern resulting from the superposition of the waves passing through the two slits.
• Compute the quantum potential based on the resulting wave function.
• Plot the interference pattern and the quantum potential.

The results are:

    """

    This code defines a double Gaussian wave function, computes its quantum potential, 

    and then plots both the wave function and the quantum potential. Adjust the parameters 

    as needed to observe different behaviors.

    """

import numpy as np

import matplotlib.pyplot as plt

def compute_quantum_potential(psi, h_bar=1, m=1):

    """

    Computes the quantum potential for a given wave function.

    Arguments:

    psi : array_like

        1-D array representing the wave function.

    h_bar : float, optional

        Value of the reduced Planck constant (default is 1).

    m : float, optional

        Mass of the particle (default is 1).

    Returns:

    quantum_potential : array_like

        1-D array representing the quantum potential.

    """

    # Compute the magnitude of the wave function

    psi_magnitude = np.abs(psi)

    # Compute the second derivative of the wave function

    second_derivative = np.gradient(np.gradient(psi_magnitude))

    # Compute the quantum potential

    quantum_potential = - (h_bar**2 / (2 * m)) * (second_derivative / psi_magnitude)

    return quantum_potential

def double_gaussian_wavefunction(x, x0, sigma, A):

    """

    Computes the double Gaussian wave function.

    Arguments:

    x : array_like

        1-D array representing the position.

    x0 : float

        Position of the center of the Gaussians.

    sigma : float

        Width of the Gaussians.

    A : float

        Amplitude of the Gaussians.

    Returns:

    psi : array_like

        1-D array representing the wave function.

    """

    psi = A * (np.exp(-((x - x0 - 1) / sigma)**2) + np.exp(-((x - x0 + 1) / sigma)**2))

    return psi

# Example usage

if __name__ == "__main__":

    # Define parameters

    x = np.linspace(-5, 5, 1000)  # Position range

    x0 = 0  # Center of the Gaussians

    sigma = 0.5  # Width of the Gaussians

    A = 1  # Amplitude of the Gaussians

    # Compute the double Gaussian wave function

    psi = double_gaussian_wavefunction(x, x0, sigma, A)

    # Compute the quantum potential

    Q = compute_quantum_potential(psi)

    # Plot the wave function and the quantum potential

    plt.figure(figsize=(12, 6))

    plt.subplot(1, 2, 1)

    plt.plot(x, psi, label='Wave function')

    plt.title('Double Gaussian Wave Function')

    plt.xlabel('Position')

    plt.ylabel('Amplitude')

    plt.legend()

    plt.subplot(1, 2, 2)

    plt.plot(x, Q, label='Quantum Potential', color='red')

    plt.title('Quantum Potential')

    plt.xlabel('Position')

    plt.ylabel('Value')

    plt.legend()

    plt.tight_layout()

    plt.show()

SBOM for AI

I think it make sense to extend to AI software assets the same approach as Embedded Systems developers are adopting for their software assets; viz. SBOM. 

Pleas see Introduction to SBOM management on embedded Linux - sergioprado.blog, and [2307.02088] Trust in Software Supply Chains: Blockchain-Enabled SBOM and the AIBOM Future (arxiv.org), and the image below for further elucidation of what SBOM is.

AI as a Tool of Evil

https://www.972mag.com/lavender-ai-israeli-army-gaza/

Wearable tech and magic carpets [Brave New World]

They also created a wireless haptic carpet that glows underfoot which not only provides a form of emergency lighting at night, but can also wirelessly transmit signals that can be used to control switches on appliances in the home, such as lights.

https://www.theguardian.com/technology/2024/apr/04/wearable-tech-how-the-human-body-can-help-power-the-future-of-smart-textiles

AI Code Generators

https://www.unite.ai/best-ai-code-generators/

And

https://www.youtube.com/watch?v=3UJ6Ei3RtKM

And

https://www.youtube.com/watch?v=2CG3P_3qJR4

نوروز 1403

 

Teach Yourself ChatGPT

I came across this no-code tutorial for a ChatGPT-based solution:

https://zapier.com/blog/custom-chatgpt/

Another way to proceed is to use a persoanl free subscription for Microsoft Azure - https://portal.azure.com/#home and look for ChatGPT.  However, that is a much more complex environmnet.

Devin AI

https://indianexpress.com/article/explained/explained-sci-tech/meet-devin-ai-first-autonomous-ai-software-engineer-9212537/

Research on Brain Responses of Programmers

From Nature:

We read, "Computer programmers show distinct, expertise-dependent brain responses to violations in form and meaning when reading code..."

Full artiche here: https://www.nature.com/articles/s41598-024-56090-6