ChatGPT ideas about amplifier design

Possibly, but the other alternative is the basis for the 1970 era Sci-Fi film (and prior book) titled, Colossus: The Forbin Project. The story Is about two defense AI computers, one American, one Russian. Which discover each other and, logically, jointly conclude and coordinate that humans are the actual threat to the planet, and determine to ’solve’ that problem. Perhaps you’ve seen it. If not, maybe check it out.
Last edited:
Agreed, there are numerous possibilities, except, one of those possibilities includes that they may logically decide to work together toward our detriment, and that we couldn’t stop them. That eventuality falls in to the quadrant of risk assessment which features a low probability of occurrence, combined with a high consequence (potentially, existential) should it occur. So, the growing fear derives from the difficult to assess potential for an extremely high consequence.

AI has the potential to not only be more intelligent than humans (not that high of a hurdle, it would seem), but so much more intelligent that they are fully under their own control, and no longer under ours. It remains a disturbingly open question as to whether they could place us under their cold, logic logic driven control. As is explored in the Sci-Fi film which I referenced.
Okay. But again, intelligence, cleverness, ingenuity, knowledge, rationality, honesty, morality, duplicity, greed, avarice, jealousy, etc., are not all equivalent nor necessarily follow from one another. To assume that high intelligence will result in everything else being ideal (from a <modern Western civilization?> human perspective) would involve a huge leap of faith.
“Prompt” in what context? I’m familiar with the use of the term in Whisper, where you feed it a sentence or two of example speech to improve audio transcription (for instance, speech interspersed with lots of “like” and “um”, etc.
You don't know what 'prompt' is within the context of ChatGPT, yet I see you post a wall of text a few posts after?

Whether an artificial construct can really 'be intelligent' or 'be conscious' or 'can create' is meaningless: you won't be able to tell the difference.

Get ready.

The prompt is you prompting it in the text area of the interface with a question. That should be evident for anyone who has done a minimal amount of looking it up.

Therefore, auto-prompting is GPT conversing with GPT back and forth.
Okay. But again, intelligence, cleverness, ingenuity, knowledge, rationality, honesty, morality, duplicity, greed, avarice, jealousy, etc., are not all equivalent nor necessarily follow from one another. To assume that high intelligence will result in everything else being ideal (from a <modern Western civilization?> human perspective) would involve a huge leap of faith.
Mark, we don’t necessarily disagree on most of your points. The main distinction is, I think, that I don’t see an assumption that all else must be ideal as necessary for a essentially worst case consequence to occur. Such consequence may only require a few of your assumptions to become true, in order to be close enough to present a seriously dangerous problem. In example, nuclear weapons, while existentially dangerous, they are within the control of humanity, as scary as is that thought. AI, however, quite conceivably might not remain within our control. As I pointed out in my prior post, as I understand it, the fear of AI derives from the potential negative consequences for humanity, should those occur. Not so much from a high probability of their occurring. Although, there does appear to be debate on exactly how low is the probability.

BTW, there is a brief, but interesting assessment of AI concerns and benefits in the April 22nd -28th issue of ‘The Economist’ magazine. Titled, “How to worry wisely about AI”. One of the worries which I had not thought about, is the potential for AI to replace many knowledge worker jobs. It used to be that workers were told that they needed to become better educated, and find a knowledge based job when they were subject to displacement by assembly automation. Now, knowledge jobs are also under threat from technology, such as ChatGPT points toward. Accountants, engineers, doctors, teachers, and many more which don’t readily come to mind. While also pointing to the positive potential of AI, human beings have to be able to earn a living to survive, or current socio-economic systems will have to radically change. I certainly don’t want societies comprised of a few ultra-rich property owners, and everyone else in squalor. One of the continuing questions of Capitalism is, what to do with human beings who are no longer needed for work, not those who are unwilling to work, but those who have no place in work.

While I don’t necessarily yet want today’s AI to autonomously engineer my next HiFi component, let alone the jet engines used on the next airliner I fly, is there any doubt, that day is coming? Perhaps, more rapidly than we currently imagine.
The G7 countries are committed to developing standards for artificial intelligence (AI) and to international discussions on how to manage it, the G7 leaders said in a statement at the Hiroshima summit.
“We are committed to further promoting multistakeholder approaches to developing standards for AI... highlighting the importance of international discussions on AI governance... while recognizing that approaches and policy instruments to achieve a shared vision and goal of robust AI can vary among different members of the G7," the final statement said.
AI can not emulate the subtleties of a sub PPM design. For now , unless I trained it (Google would violate my IP) and read my threads.
AI can't do this (below) .... yet ?? . 10 years of making perfect A/B amps for my fellow organics. ;)

I do see a corporate race to exploitation corrupting any AI that becomes available. Just look at the internet as a whole ??
Mr. Altman testifies to Congress Idiots who are so far behind .... Whoa !!


  • mini-ops.JPG
    165.1 KB · Views: 54
Only that version doesn't but it's already superseded many times over in power by newer systems.

It's all going to Natural Language automation of applications. Give it simple commands to do complex stuff and it will execute faster and better than humans.

If self-prompting is a chain of thoughts, the next iterations using Natural Language Decision Trees with auto-evaluation are going to be crazy.
And that's only the beginning. Look at what the internet is. AI will turn in to the ULTIMATE CrapHause.
If one of the innovators of AI testifies , even before AI's ultimate effect ...that it might make for trouble ???
IT IS EVIL. If the creators warn of it , even before it is mainstream ... we should watch out !!

  • Like
Reactions: 1 user
for audio amplifiers what factors affect whether to use SM components over TH components

When considering whether to use surface mount (SM) components or through-hole (TH) components in audio amplifiers, several factors come into play. The choice between these two types of components depends on various considerations, including:

Size and Space Constraints: SM components are significantly smaller and occupy less space on the PCB compared to TH components. If the audio amplifier design requires compactness or is intended for portable devices with limited space, SM components are often preferred.

Signal Integrity and Noise: TH components generally offer better signal integrity due to their larger lead structures, which can provide better electrical connections and reduce noise and interference. In some cases, the larger size of TH components can lead to improved performance in terms of minimizing signal degradation.

Power Handling: TH components typically have higher power handling capabilities compared to their SM counterparts. If the audio amplifier requires high power output or needs to handle significant currents, TH components may be more suitable to withstand the associated power demands.

Manufacturing Process: SM components are typically more suited for automated manufacturing processes such as surface mount technology (SMT), which enables faster and more cost-effective assembly. TH components, on the other hand, often require manual insertion and soldering, which can increase production time and cost.

Availability and Cost: SM components are widely used in modern electronics, making them more readily available and often less expensive than TH components. However, some specialized or high-power components may still be primarily available in through-hole packages.

Repair and Maintenance: If repairability or ease of replacement is a consideration, TH components can be advantageous. They are generally easier to remove and replace manually, making troubleshooting and maintenance more straightforward.

It's important to note that the choice between SM and TH components is ultimately dependent on the specific requirements and constraints of the audio amplifier design. Engineers often consider a combination of these factors along with their experience and the overall system objectives when selecting the appropriate components.

would complicated topologies do better with SM components or TH components, in terms of footprint how do capacitors, resistors, semiconductors fare

In complicated topologies, the choice between surface mount (SM) components and through-hole (TH) components depends on various factors, including the complexity of the circuit, the size constraints, and the specific components involved. Here's a breakdown of how different types of components fare in terms of footprint:

Capacitors: Surface mount capacitors are available in a wide range of sizes, including small footprints such as 0402, 0603, and 0805, which are commonly used in compact circuits. They offer high capacitance values and are suitable for applications where space is limited. However, for larger capacitance values or high-voltage applications, through-hole capacitors might be preferred due to their larger physical size and better power handling capabilities.

Resistors: Like capacitors, surface mount resistors are available in various sizes, including small form factors suitable for compact designs. SM resistors are commonly used in modern electronics due to their small size, accuracy, and availability. Through-hole resistors are typically larger and can handle higher power levels, making them suitable for high-power applications or environments where heat dissipation is a concern.

Semiconductors: The footprint of semiconductors, such as integrated circuits (ICs), depends on the specific package type used. Surface mount packages, such as small outline integrated circuits (SOIC) or quad flat no-leads (QFN), are commonly used for compact and densely populated circuit boards. These SM packages offer smaller footprints and better thermal performance. However, through-hole packages, such as dual in-line packages (DIP), may still be preferred for certain specialized components or when ease of manual replacement is a priority.

In general, complicated topologies can benefit from the use of surface mount components due to their smaller footprints, allowing for denser circuit layouts and improved signal routing. SM components are often chosen for their miniaturization, automated assembly advantages, and availability. However, in situations where high power handling or specific component requirements come into play, through-hole components might be necessary, even if they occupy larger footprints. The selection of components ultimately depends on the specific needs and constraints of the circuit design.