# The potential of artificial general intelligence tools like ChatGPT for design facilitation

Artificial general intelligence (AGI) is here, and it is changing the way we live and work. This article highlights the potential of AGI as a powerful design facilitation tool and aims to dispel the misconception that AGI is merely an elaborate encyclopedic network that recites answers from a database. While much has been written about AGI from the perspectives of the humanities and social sciences, this article takes a STEM-based approach to demonstrate how AGI can be incorporated as an ally in design strategies.

ChatGPT, a technology developed by OpenAI and currently at Version 4.0, is a prime example of AGI in action [1]. Preliminary reports suggest that this new technology is already threatening up to 80% of skilled jobs in the United States alone, spanning a wide section of the humanities and social sciences from writers to editors to advisors and many more [2]. However, the potential of ChatGPT extends far beyond its ability to automate tasks traditionally performed by humans. By asking the right questions, we can unlock the true power of AGI as a design facilitation tool.

This article demonstrates how AGI can be incorporated into the design process to drive innovation and progress. Through a series of simple questions posed to ChatGPT, we show that AGI has the potential to transform the design process by generating new ideas, identifying potential flaws, and providing real-time feedback. By embracing AGI as a powerful tool for innovation and progress, we can stay ahead of the curve and remain competitive in an ever-changing world. It is time to ask the right questions and embrace AGI as a powerful tool for innovation and progress.

In the 2004 film “I, Robot”, loosely based on Isaac Asimov’s 1950 novel of the same name, the protagonist, played by Will Smith, engages in a series of exchanges with the virtual AI interface, Dr. Alfred Lanning. For each misguided question he asks, he receives the response, “I’m sorry, my responses are limited, you must ask the right question.” Only when he finally asks the right question does he receive the answer, “That is the right question.” This scenario is reminiscent of interactions with ChatGPT. Preliminary assessments of this technology have mainly come from non-technical fields, leading some to belittle it as an elaborate encyclopedic network that simply reads and recites answers from a database. However, in this DI, we aim to show that this sentiment is not entirely accurate.

To maintain concision, we present ChatGPT with four questions covering key aspects of engineering, including mechanics, mathematics, and electronics. These questions are listed below together with the ChatGPT’s responses. The questions themselves are kept simple. They are framed as sentences that have a mixture of numerical and worded values in an attempt to confound ChatGPT. Furthermore, note that the responses are copied as-is from the ChatGPT terminal. The reader will note that the output formatting is mine, but identical in content to ChatGPT.

**Question 1:**

**Response 1:**

F_net = m*a

where F_net is the net force acting on the object, m is its mass, and a is its acceleration.

First, we need to calculate the gravitational force acting on each object. The gravitational force is given by:

F_gravity = m*g

where m is the mass of the object and g is the acceleration due to gravity (9.8 m/s^2).

For the 2 kg block on the incline, the gravitational force acting on it is:

F_gravity1 = m1gcos(30) = 29.8cos(30) = 16.98 N

For the 4 kg weight hanging freely, the gravitational force acting on it is:

F_gravity2 = m2g = 49.8 = 39.2 N

Next, we need to calculate the tension force in the string connecting the two objects. Since the string is rigid and does not stretch, the tension force in the string is the same throughout its length. We can calculate this tension force using the net force acting on the 4 kg weight:

F_net = F_gravity2 – T = m2*a

where T is the tension force in the string and a is the acceleration of the system.

Substituting the values we know, we get:

T = F_gravity2 – m2*a

Now, we can use the force of gravity acting on the 2 kg block and the tension force in the string to calculate the net force acting on the block:

F_net = T – F_gravity1 = m1*a

Substituting the values we know, we get:

T – F_gravity1 = m1*a

T = 39.2 – 2*a

Substituting this expression for T into the previous equation, we get:

(39.2 – 2a) – F_gravity1 = m1a

Simplifying and solving for a, we get:

a = (39.2 – F_gravity1)/(m1 + m2)

a = (39.2 – 16.98)/(2 + 4)

a = 4.04 m/s^2

Therefore, the acceleration of the 2 kg block is 4.04 m/s^2.

**Question 2:**

**Response 2:**

However, there seems to be an issue with the sixth pair as it contains non-numeric values. Assuming it to be a typo and correcting it to (6,7.5), we can proceed with the linearization process.

Here’s the table of the given data:

Experiment Observation

1 3

2 4.2

3 4.3

4 5.1

6 7.5

7 8.9

And here’s a plot of the data:

Experiment Pairs Plot

From the plot, we can see that the data points appear to form a roughly linear relationship. We can calculate the slope and intercept of the line of best fit using linear regression.

Using a spreadsheet program or an online tool, we can calculate the slope and intercept to obtain the linear equation:

y = 0.7029x + 2.0729

where y is the observation and x is the experiment number.

Therefore, the linearized experiment pairs are:

Experiment Observation

1 2.7758

2 3.4783

3 4.181

4 4.8835

6 6.5895

7 7.292

Note that the linearized observation values are not exactly the same as the original observation values. They represent the expected observation values based on the linear relationship between the experiment number and observation.

Comment: While ChatGPT’s ability to decipher the input is remarkable, the final answer is INCORRECT. The correct answer is y = 0.9503x + 1.8571.

What went wrong? It is likely ChatGPT’s reliance on some algorithm from an obscure source.

**Question 3:**

**Response 3:**

The impedance of a capacitor is given by:

Zc = 1/(jwC)

where j is the imaginary unit, w is the angular frequency (2pif), and C is the capacitance in farads.

In this case, the capacitance is 10 microfarads, or 1010^-6 farads, and the frequency is 10 kilohertz, or 1010^3 Hz. Therefore, the angular frequency is:

w = 2pi10^3 = 6.28*10^4 rad/s

The impedance of the capacitor is:

Zc = 1/(j6.2810^41010^-6) = -j25.32 ohms

The impedance of a resistor is simply its resistance. In this case, the resistance is 1 kiloohm.

The total impedance of the circuit is the sum of the impedance of the resistor and the impedance of the capacitor:

Ztotal = R + Zc = 1 kiloohm – j25.32 ohms

To determine if the circuit is resonant, we need to calculate the frequency at which the impedance of the circuit is purely resistive (i.e., the imaginary part of the impedance is zero). This occurs at the resonant frequency, which is given by:

f = 1/(2piR*C)

Substituting the values we know, we get:

f = 1/(2pi110^310*10^-6) = 15.92 kilohertz

Since the given frequency of 10 kilohertz is less than the resonant frequency of 15.92 kilohertz, the circuit is not resonant. The impedance of the circuit is a complex number with a nonzero imaginary part, indicating that there is a phase shift between the voltage and current in the circuit.

**Question 4:**

**Response 4:**

The impedance of a capacitor is given by:

Z_c = 1/(jωC)

where j is the imaginary unit, ω is the angular frequency (2πf), and C is the capacitance in farads.

At a frequency of 10 kHz, the angular frequency is:

ω = 2πf = 2π × 10,000 = 62,831.85 rad/s

The impedance of the capacitor is then:

Z_c = 1/(j × 62,831.85 × 10^-6) = -j15.92 ohms

The impedance of the resistor is simply its resistance, which is 1 kiloohm.

The total impedance of the series circuit is the sum of the impedance of the capacitor and the resistor:

Z_total = Z_c + R = -j15.92 + 1000 = 1000 – j15.92 ohms

The magnitude of the voltage across the capacitor can be found using Ohm’s law:

V_c = I × Z_c

where I is the current flowing through the circuit.

The current can be found using the voltage and total impedance:

I = V / Z_total

where V is the voltage applied to the circuit, which is 1 volt.

I = 1 / (1000 – j15.92) = 0.999 – j0.016 A

Finally, the voltage across the capacitor is:

V_c = I × Z_c = (0.999 – j0.016) × (-j15.92) = 0.253 + j15.891 mV

The magnitude of the voltage across the capacitor is therefore:

|V_c| = √(0.253^2 + 15.891^2) = 15.894 mV (rounded to 3 decimal places)

So the magnitude of the voltage across the capacitor is approximately 15.894 mV.

ChatGPT’s ability to generate new knowledge through unsupervised learning is unprecedented. However, it is clear that this AGI language model still requires human input to provide the necessary guidance and focus. It is uncertain how long this will remain the case, as AGI continues to evolve into something more powerful. The potential for AGI to problem solve and generate new knowledge independently can make it an extremely useful tool in engineering repertoire at any point in the design cycle.

*Professor Ocaya specializes in electronics and solid-state physics, which he teaches at the Qwaqwa Campus of the UFS. He is active in computing, mathematical methods, new techniques for device characterization, material science and microcontroller-based instrument design. He holds a C3 rating from the National Research Foundation (NRF) of South Africa.*

**Related Content**

Changing trends in designing space electronics

2023: A technology forecast for the year ahead

Multi-tier machine learning helps speed edge AI deployment

AI-powered EDA tool suite assists chipmakers

4 basic steps in implementing an AI-driven design workflow

**References**

ChatGPT, https://chat.openai.com/chat

Motherboard: Tech by Vice, OpenAI Research Says 80% of U.S. Workers’ Jobs Will Be Impacted by GPT, https://www.vice.com/en/article/g5ypy4/openai-research-says-80-of-us-workers-will-have-jobs-impacted-by-gpt

googletag.cmd.push(function() { googletag.display(‘div-gpt-ad-native’); });

–>

The post The potential of artificial general intelligence tools like ChatGPT for design facilitation appeared first on EDN.

## Leave a Reply

Want to join the discussion?Feel free to contribute!