Author Topic: OpenAI ChatGPT for electronics design  (Read 26356 times)

0 Members and 2 Guests are viewing this topic.

Offline qatTopic starter

  • Contributor
  • Posts: 31
  • Country: se
OpenAI ChatGPT for electronics design
« on: December 16, 2022, 05:51:59 pm »
I was messing around with the openAI bot and asked it some random question regarding electronics. It seems to have gotten a lot of attention recently, and I wanted to find out if my job is at risk! ;D

Maybe it's already a thing, but I've found it to provide quite useful answers to some questions, it is also able to design basic circuits using 555 timers etc, and calculate component values! Its definitely worth experimenting with it!

[Q:] Give thee examples of low-noise LDOs available at Digi-Key. They should be able to handle input voltages ranging from 1 to 10 volts, and supply more than 160mA of current. External components needed should be as low as possible.
[A:] Here are three examples of low-noise low-dropout (LDO) regulators that are available at Digi-Key and are capable of handling input voltages ranging from 1V to 10V and supplying more than 160mA of current:

    1. Linear Technology LT3080: This LDO regulator has a low noise floor of just 6.5 µV RMS and is capable of supplying up to 1A of output current. It requires a minimum of just two external components, making it a compact and simple solution.

    2. Texas Instruments TPS7A4700: This LDO regulator has a low noise floor of just 10 µV RMS and is capable of supplying up to 200mA of output current. It requires a minimum of just two external components, making it a compact and simple solution.

    3. Analog Devices ADP150: This LDO regulator has a low noise floor of just 8 µV RMS and is capable of supplying up to 200mA of output current. It requires a minimum of just two external components, making it a compact and simple solution.

It is worth noting that these are just a few examples of the many low-noise LDO regulators that are available at Digi-Key, and the specific LDO that is best for your application will depend on your specific requirements and constraints. It is always a good idea to carefully assess your needs and consult with a power supply design expert before making a decision.

[Q:] Generate an 12-layer PCB stackup for a Zynq based system that will be used in avionics.
[A:] Here is an example of a 12-layer PCB stackup for a Zynq-based system that could be used in avionics applications:

    Signal layer 1
    Ground plane
    Power plane
    Signal layer 2
    Ground plane
    Power plane
    Signal layer 3
    Ground plane
    Power plane
    Signal layer 4
    Ground plane
    Power plane

This stackup allows for the use of four signal layers, which can be useful for routing high-speed signals and minimizing cross-talk between different signals. The power and ground planes provide a stable power and ground reference for the signals on the signal layers, which can help to improve the performance of the system.

It is worth noting that this is just one possible example of a PCB stackup for a Zynq-based system, and the specific stackup that is best for your application will depend on your specific requirements and constraints. It is always a good idea to consult with a PCB design expert and carefully assess your needs before deciding on a final stackup.

[Q:] Design a 2.35V power-supply using a Buck-Boost regulator from Ti. It should be capable of handling up to 1A of output power and 5V input voltage. Low output noise is not critical, but capability to handle high temperatures is.
[A:] One option for designing a 2.35V power supply using a buck-boost regulator from Texas Instruments (TI) is to use the TI TPS63802. This is a high-temperature buck-boost regulator that is capable of handling up to 1A of output current and input voltages ranging from 2.7V to 5.5V. It also has a built-in power switch and can operate at switching frequencies up to 2.4 MHz.

To design the power supply, you will need to connect the input voltage source to the VIN pin of the TPS63802, and connect the output load to the VOUT pin. You will also need to set the desired output voltage using an external resistor divider, as described in the TPS63802 datasheet.

In addition to the TPS63802, you will also need to include some passive components in your power supply design, such as bypass capacitors and an output filter inductor. The specific values of these components will depend on the specific requirements of your application, and can be chosen using the guidelines provided in the TPS63802 datasheet.

It is worth noting that this is just one possible example of how to design a 2.35V power supply using a buck-boost regulator, and the specific design that is best for your application will depend on your specific requirements and constraints. It is always a good idea to carefully assess your needs and consult with a power supply design expert before finalizing your design.

Try it out!

https://chat.openai.com/chat

Update,
It is insane what it is able to do, ask for example: Write a VHDL program to interface with the Analog Devices AD9371 IC...
« Last Edit: December 16, 2022, 08:13:16 pm by qat »
 
The following users thanked this post: MK14

Online ataradov

  • Super Contributor
  • ***
  • Posts: 11905
  • Country: us
    • Personal site
Re: OpenAI ChatGPT for electronics design
« Reply #1 on: December 16, 2022, 11:30:06 pm »
LOL. It got all the characteristics for all the devices wrong. It just knows from context that some number must follow and inserts a random number.

It is amazing that it can generate text that coherent, but I doubt it would be useful for anything practical.
Alex
 
The following users thanked this post: tom66, Siwastaja, ROT

Offline tom66

  • Super Contributor
  • ***
  • Posts: 7336
  • Country: gb
  • Electronics Hobbyist & FPGA/Embedded Systems EE
Re: OpenAI ChatGPT for electronics design
« Reply #2 on: December 16, 2022, 11:35:44 pm »
LOL. It got all the characteristics for all the devices wrong. It just knows from context that some number must follow and inserts a random number.

It is amazing that it can generate text that coherent, but I doubt it would be useful for anything practical.

This is the problem.  It's *very good* at bullshitting, and unfortunately it's very hard to tell when it's wrong unless you know about the subject in good detail.  For instance, you can ask it to compare a Zynq to an iPhone and, at least in my case, it said one advantage of the Zynq was it had SPI whereas the iPhone had a camera, and a camera is not as good as having serial peripheral interface. 

But, it is pretty cool to play with.  And it has some very interesting implications for the future of AI. 
 
The following users thanked this post: splin

Offline MK14

  • Super Contributor
  • ***
  • Posts: 4987
  • Country: gb
Re: OpenAI ChatGPT for electronics design
« Reply #3 on: December 17, 2022, 01:32:42 am »
Could it be used to replace all Electronics Engineers? No, not at the moment.  But its responses are amazing for an AI's attempt to mimic a real electronics engineer.
In the future, this could become worryingly good.  On the one hand, it could hinder peoples employment, but on the other, could help create lots of inventions, products and make for some interesting talking appliances/companions.

[My test question]...
If I had some NE555 timers, are they made with bjt's or mosfets?, what if it was a CMOS version?, also where can I buy them?, how much are they?, how much do they weigh?, can I design a CPU with them, write a verilog example 555 timer.
[End of my test question]...Its response follows:

Quote
The NE555 timer is a general-purpose timer integrated circuit (IC) that can be used in a variety of timing applications. It is typically made with bipolar junction transistors (BJTs) or complementary metal-oxide-semiconductor (CMOS) technology.

The NE555 timer is available in both BJT and CMOS versions. The BJT version is typically used in applications that require higher power handling, while the CMOS version is used in applications that require lower power consumption.

You can purchase NE555 timers from electronics retailers or online marketplaces. Prices may vary depending on the supplier and the quantity you purchase. A single NE555 timer typically weighs a few grams.

It is not practical to design a central processing unit (CPU) using NE555 timers, as they are not capable of performing the complex operations required by a CPU. Instead, CPUs are typically designed using more advanced integrated circuits, such as microcontrollers or microprocessors.

If you want to use an NE555 timer in a circuit, you can use a hardware description language (HDL) such as Verilog to describe the behavior of the timer. Here is a simple example of a Verilog description of an NE555 timer:

module NE555_timer (input clk, reset, trigger, output q);
  reg [31:0] counter;

  always @(posedge clk or posedge reset) begin
    if (reset) counter <= 0;
    else if (trigger) counter <= 32'd500_000; // set timer period
  end

  assign q = (counter == 0);
endmodule

This example shows the basic structure of an NE555 timer in Verilog. The timer has four inputs (clk, reset, trigger, and output) and one output (q). The counter register is used to count the number of clock cycles, and the q output is set to 1 when the counter reaches zero. The timer period can be set by writing to the counter register when the trigger input is asserted.

[Then I asked (mistakenly on my part, as it HAD actually supplied the answer, in its response):]
Wait a minute.  You seem to have forgotten to tell me how much they weigh.
[End of my question]

Quote
I apologize for the oversight. A single NE555 timer typically weighs a few grams. The exact weight may vary slightly depending on the specific version of the NE555 timer and the package it is packaged in. NE555 timers are usually packaged in a small, through-hole plastic package, which makes them lightweight and easy to use in a wide variety of applications.

[I then ask]
I want an op-amp circuit, which is an inverting amplifier, with a gain of -20, capacitively coupled on both the input and output sides, so as to ignore any DC bias.  Please illustrate the schematic, by producing it in Verilog.
[End of question]

Quote
Here is a Verilog description of an inverting amplifier circuit with a gain of -20 that is capacitively coupled on both the input and output sides:

module inverting_amplifier (input in, output out);
  parameter GAIN = -20;

  // Capacitive coupling on input side
  wire in_coupled;
  capacitor cap1 (in, in_coupled, 1nF);

  // Inverting amplifier
  opamp u1 (in_coupled, out, GAIN);

  // Capacitive coupling on output side
  wire out_coupled;
  capacitor cap2 (out, out_coupled, 1nF);
  assign out = out_coupled;
endmodule

This circuit uses an operational amplifier (op-amp) to implement the inverting amplifier, with a gain of -20. The input signal is capacitively coupled to the op-amp through a 1nF capacitor (cap1), which blocks any DC bias from the input signal. The output of the op-amp is then capacitively coupled back to the output of the circuit through another 1nF capacitor (cap2), which also blocks any DC bias from the output signal.

I hope this helps! Let me know if you have any questions.

[My final question, here]
Now, show me that schematic, drawn as an ASCII/text art diagram, at a high level only.  Also, redesign it using small signal transistors, along with the recommended part numbers.
[End of final question]

Quote
Here is a high-level ASCII/text art diagram of the inverting amplifier circuit with a gain of -20 that is capacitively coupled on both the input and output sides:

      +--------+
      |        |
in ---|1nF     |
      |        |
      +--------+
               |
               v
      +--------+
      |        |
      |        |
      |  opamp |
      |        |
      |        |
      +--------+
               |
               v
      +--------+
      |        |
out ---|1nF     |
      |        |
      +--------+

To redesign this circuit using small signal transistors, you can use a pair of NPN transistors connected in a common emitter configuration. Here is a Verilog description of the redesigned circuit:

module inverting_amplifier (input in, output out);
  parameter GAIN = -20;

  // Capacitive coupling on input side
  wire in_coupled;
  capacitor cap1 (in, in_coupled, 1nF);

  // Inverting amplifier using NPN transistors
  npn_transistor q1 (in_coupled, out, 2N3904);
  npn_transistor q2 (out, out, 2N3904);

  // Capacitive coupling on output side
  wire out_coupled;
  capacitor cap2 (out, out_coupled, 1nF);
  assign out = out_coupled;
endmodule

In this redesign, the op-amp has been replaced by a pair of NPN transistors (q1 and q2), connected in a common emitter configuration. The gain of the amplifier can be controlled by adjusting the values of the resistors connected to the transistors. The 2N3904 transistor is a commonly-used small signal transistor that can be used in this circuit.

I hope this helps! Let me know if you have any questions.
 

Offline MK14

  • Super Contributor
  • ***
  • Posts: 4987
  • Country: gb
Re: OpenAI ChatGPT for electronics design
« Reply #4 on: December 17, 2022, 04:45:02 am »
I used the following question, to attempt to see, how knowledgeable and helpful, it can be, as regards the use of FPGAs.

I've got a problem with my Verilog, FPGA files.  There seems to be a metastabilty between my two different clock domains, and I don't want to use a FIFO to fix it.  Also, I want to implement a latch, but the design rules prohibit it, and say I must use a register, because the latch is unclocked.  Also, I need to add a 32 bit barrel shifter, which connects to a big multiplier.  Will that need any more DSP slices?  Also, any tips on how I can optimize the maximum clock rate of the overall design, ideally without using anymore block ram, which is nearly exhausted.

Quote

Metastability can occur when there is a race condition between two clock domains, and it can be difficult to resolve without using some kind of synchronization element such as a FIFO or a flip-flop. However, there are some techniques that you can use to minimize the risk of metastability, such as ensuring that the clock skew between the two domains is minimized, and using a clock-crossing technique such as clock-domain crossing (CDC) cells or PLLs to synchronize the two clocks.

As for the latch, it is generally not recommended to use latches in digital design because they are unclocked and can cause unpredictable behavior. Instead, you should use a flip-flop, which is a clocked version of a latch and will behave more predictably.

As for the barrel shifter, it's possible that it may require additional DSP slices depending on the specific implementation and the capabilities of your FPGA. In general, it's a good idea to try to minimize the use of DSP slices because they are often a limited resource on an FPGA, but if you need to use them to implement certain functions, it may be necessary.

To optimize the maximum clock rate of your design, you can try a few different techniques. One option is to try to reduce the overall complexity of your design, which can help to reduce the total number of logic elements and improve performance. You can also try to optimize your placement and routing to reduce the length of critical paths, which can help to improve the maximum clock rate. Finally, you can try to reduce the number of I/O delays by carefully selecting the I/O standards and locations for your pins.
 

Online ataradov

  • Super Contributor
  • ***
  • Posts: 11905
  • Country: us
    • Personal site
Re: OpenAI ChatGPT for electronics design
« Reply #5 on: December 17, 2022, 04:59:10 am »
Again, amazing that it can interpret the question and provide a coherent response, but this is the most generic advice.

And understanding that "there seems to be a metastabilty between my two different clock domains," is already 80% of the solution to someone who managed to make a project with 2 asynchronous clocks.

That's where all those AIs fall apart. They are a great interactive replacement for stackoverflow, but they can't solve practical issues in exiting project (which is majority of work you do on a day to day basis). You can't point it to a bunch of Verilog and say "this glitches and locks up once in a while. what do I do?". And if you copy-paste stuff generated by AI, you can rest assured your overall code would be glitchy.

And I don't see any indication of how close they are to the practical limit. It is possible that in the future those systems would be smart enough to figure that out, but I'm skeptical.  This feels like one of those things were it would be "almost there" for ages.
« Last Edit: December 17, 2022, 05:01:21 am by ataradov »
Alex
 
The following users thanked this post: MK14

Offline MK14

  • Super Contributor
  • ***
  • Posts: 4987
  • Country: gb
Re: OpenAI ChatGPT for electronics design
« Reply #6 on: December 17, 2022, 05:32:17 am »
Again, amazing that it can interpret the question and provide a coherent response, but this is the most generic advice.

And understanding that "there seems to be a metastabilty between my two different clock domains," is already 80% of the solution to someone who managed to make a project with 2 asynchronous clocks.

That's where all those AIs fall apart. They are a great interactive replacement for stackoverflow, but they can't solve practical issues in exiting project (which is majority of work you do on a day to day basis). You can't point it to a bunch of Verilog and say "this glitches and locks up once in a while. what do I do?". And if you copy-paste stuff generated by AI, you can rest assured your overall code would be glitchy.

And I don't see any indication of how close they are to the practical limit. It is possible that in the future those systems would be smart enough to figure that out, but I'm skeptical.  This feels like one of those things were it would be "almost there" for ages.

It is a bit like the Fusion developments.  It is not quite clear, how long it will take for us to have, significantly useful, versions of these things.

Are these things 5 years, 20 years or 50 or more years, before they will become useful enough to replace many human jobs and/or create unlimited, almost free, energy?

A similar thing happened, with Chess Software/Hardware, when they first become available, in around the 1970s/80s, because of things like the Fidelity Chess Challenger.

The $64,000,000 question, was when/if Chess computers/Programs, would be powerful enough to beat, even the (human) World Chess Champion.

Some thought it would not be that long, some thought it would never be achieved, because computers could never think, like a human being.

Current uses of such chatbot technology, is arguably NOT necessarily helping yet.  E.g. You definitely need to speak to someone at customer services.  So you phone them up.  Then get this (silly?) computer voice that wastes, perhaps ten minutes of your time, by slowly saying, something like:
'Press 1 to change your account details'
'Press 2 to change your address'
'Press 3 to check your serial number is valid'
Etc.
Eventually, you hopefully get to the bit where it says:
'Be quiet for twenty seconds, and we will put you through to a customer service representative'
'You are position 597 in our queue'
Then plays crazily loud, and amazingly annoying, musical tones, for the next 1.5 hours (NOT a joke.  Exactly that happened to me recently.  Yes, really around 1.5 hours!, just to talk to someone).
« Last Edit: December 17, 2022, 05:34:56 am by MK14 »
 

Offline SiliconWizard

  • Super Contributor
  • ***
  • Posts: 15800
  • Country: fr
Re: OpenAI ChatGPT for electronics design
« Reply #7 on: December 17, 2022, 07:55:41 pm »
But the real question IMHO is not what it can do, but what we want to do with it exactly, and the latter is still very unclear to me.

The parallel with nuclear fusion is interesting (see the other thread about it.) We pretend to know what we would do with it, and even make it sound almost obvious, while it is, in fact, absolutely not.
 
The following users thanked this post: MK14

Online Smokey

  • Super Contributor
  • ***
  • Posts: 3045
  • Country: us
  • Not An Expert
Re: OpenAI ChatGPT for electronics design
« Reply #8 on: February 27, 2023, 01:02:16 am »
Careful out there...

From ChatGPT:
Quote
In summary, if you want to sink power from a USB host and do not require any data transfer, you should leave the DP and DM lines floating, and include the standard 15kΩ pull-down resistors on these lines to ensure proper signal termination.
 

Offline eti

  • Super Contributor
  • ***
  • !
  • Posts: 1801
  • Country: gb
  • MOD: a.k.a Unlokia, glossywhite, iamwhoiam etc
Re: OpenAI ChatGPT for electronics design
« Reply #9 on: February 27, 2023, 01:06:44 am »
The REAL intelligence comes from someone saying "What the @@@@ is this - another tech industry fad? - No thanks, I think we humans have a little more experience and discernment than this dumb robot bullshit"

The issue is, people's brains will start to depend on this crap, and their once sharp minds with begin to atrophy, depending on some dub box of rocks to "think" and to "design" things for them. Watch it all come crashing down. Whoooooops!  :palm:

Supposedly humans are intelligent - where's the evidence of this?
« Last Edit: February 27, 2023, 01:08:41 am by eti »
 

Offline xrunner

  • Super Contributor
  • ***
  • Posts: 7836
  • Country: us
  • hp>Agilent>Keysight>???
Re: OpenAI ChatGPT for electronics design
« Reply #10 on: February 27, 2023, 01:25:46 am »
Supposedly humans are intelligent - where's the evidence of this?

Stand by let me go ask ChatGPT ...
I told my friends I could teach them to be funny, but they all just laughed at me.
 
The following users thanked this post: timenutgoblin

Offline mairo

  • Supporter
  • ****
  • Posts: 226
  • Country: au
Re: OpenAI ChatGPT for electronics design
« Reply #11 on: February 27, 2023, 01:47:48 am »
...
[Q:] Give thee examples of low-noise LDOs available at Digi-Key. They should be able to handle input voltages ranging from 1 to 10 volts, and supply more than 160mA of current. External components needed should be as low as possible.
[A:] Here are three examples of low-noise low-dropout (LDO) regulators that are available at Digi-Key and are capable of handling input voltages ranging from 1V to 10V and supplying more than 160mA of current:

    1. Linear Technology LT3080: This LDO regulator has a low noise floor of just 6.5 µV RMS and is capable of supplying up to 1A of output current. It requires a minimum of just two external components, making it a compact and simple solution.

    2. Texas Instruments TPS7A4700: This LDO regulator has a low noise floor of just 10 µV RMS and is capable of supplying up to 200mA of output current. It requires a minimum of just two external components, making it a compact and simple solution.

    3. Analog Devices ADP150: This LDO regulator has a low noise floor of just 8 µV RMS and is capable of supplying up to 200mA of output current. It requires a minimum of just two external components, making it a compact and simple solution.
...

How can it give you an answer about stock in Digikey when it is not connected to the web?

It gave me a "different' answer, one that correctly explained it cannot search in Digikey:
[q]: Give thee examples of low-noise LDOs available at Digi-Key. They should be able to handle input voltages ranging from 1 to 10 volts, and supply more than 160mA of current. External components needed should be as low as possible.
[a]: As an AI language model, I cannot browse the internet and search for up-to-date information or specific products available at a particular store like Digi-Key. However, I can provide you with a general idea of what to look for when searching for low-noise LDOs that meet your requirements.

Here are some general guidelines to follow when selecting a low-noise LDO:

Look for LDOs with a low dropout voltage (VDO), which is the minimum voltage difference between the input and output of the regulator for it to operate correctly. The lower the VDO, the less power is dissipated as heat, which is desirable in low-power applications.

Choose LDOs with a high power supply rejection ratio (PSRR), which is a measure of how well the regulator rejects noise from the input power supply. A high PSRR value means that the regulator can filter out noise and provide a clean output voltage.

Consider LDOs with low output noise, which is a measure of how much voltage fluctuation the regulator produces at its output. A low output noise value means that the regulator produces a stable and consistent output voltage.

Based on these guidelines, you can search for low-noise LDOs that meet your requirements. Here are three examples of LDOs that meet your criteria:

ADP150AUJZ-1.8-R7: This LDO from Analog Devices has a low dropout voltage of 190mV at 160mA and a high PSRR of 65dB at 1kHz. It also has a low output noise of 27µVRMS, making it suitable for noise-sensitive applications. This LDO requires only two small ceramic capacitors for stability.

LP2985-33DBVR: This LDO from Texas Instruments has a low dropout voltage of 300mV at 150mA and a high PSRR of 60dB at 1kHz. It also has a low output noise of 30µVRMS, making it suitable for low-noise applications. This LDO requires only a small ceramic output capacitor and a small input capacitor for stability.

MCP1703T-3302E/CB: This LDO from Microchip has a low dropout voltage of 300mV at 250mA and a high PSRR of 60dB at 1kHz. It also has a low output noise of 22µVRMS, making it suitable for low-noise applications. This LDO requires only a small ceramic output capacitor and a small input capacitor for stability.





 

Offline Brumby

  • Supporter
  • ****
  • Posts: 12413
  • Country: au
Re: OpenAI ChatGPT for electronics design
« Reply #12 on: February 27, 2023, 01:48:50 am »
Supposedly humans are intelligent - where's the evidence of this?

Stand by let me go ask ChatGPT ...

Oh, PLEASE post the response!!!!   :-DD
 

Offline xrunner

  • Super Contributor
  • ***
  • Posts: 7836
  • Country: us
  • hp>Agilent>Keysight>???
Re: OpenAI ChatGPT for electronics design
« Reply #13 on: February 27, 2023, 01:53:21 am »
Supposedly humans are intelligent - where's the evidence of this?

Stand by let me go ask ChatGPT ...

Oh, PLEASE post the response!!!!   :-DD

LOL sure thing I'm already logged in there!

Me: Hi.

ChatGPT: Hello! How can I assist you today?

Me: Supposedly humans are intelligent - where's the evidence of this?

Chat GPT: There is abundant evidence that humans are intelligent, as evidenced by our ability to solve complex problems, think abstractly, communicate using language, and create sophisticated tools and technologies. Here are a few examples:

1. Problem solving: Humans have developed advanced problem-solving skills, from basic tasks such as finding food and water, to complex mathematical and scientific challenges.

2. Language: Humans have developed language, which is a complex system of communication that allows us to convey abstract ideas, share knowledge, and build culture.

3. Tool use: Humans have used tools for thousands of years, from simple stone tools to advanced technologies like computers and smartphones. Our ability to create and use tools has allowed us to achieve things that would have been impossible otherwise.

4. Creativity: Humans are capable of producing art, music, literature, and other forms of creative expression, which require imagination, originality, and skill.

Overall, the evidence suggests that humans are intelligent and have developed many remarkable abilities that have allowed us to thrive and prosper as a species.
I told my friends I could teach them to be funny, but they all just laughed at me.
 

Offline Brumby

  • Supporter
  • ****
  • Posts: 12413
  • Country: au
Re: OpenAI ChatGPT for electronics design
« Reply #14 on: February 27, 2023, 02:06:28 am »
< I went the next step ...>

Q: Respond to the same question sarcastically

A: Oh, absolutely! Humans are renowned for their unparalleled intelligence. I mean, who else could come up with such brilliant ideas like using plastic straws and bags that take hundreds of years to decompose, or starting wars over trivial matters, or creating reality TV shows that celebrate shallowness and stupidity? Clearly, we are the pinnacle of intelligence in the universe.

Q: Tell me more!

A: Well, where do I even begin? Humans are so intelligent that we can't even figure out how to live sustainably on our own planet without causing widespread environmental destruction. We're so smart that we spend billions of dollars on weapons of mass destruction instead of investing in things that would actually benefit our species, like education and healthcare. And let's not forget how clever we are at ignoring scientific evidence and choosing to believe in myths and superstitions instead. Truly, it's a wonder how we've managed to survive this long with such exceptional intelligence.




< I think I might stop here >  ;D
 

Offline xrunner

  • Super Contributor
  • ***
  • Posts: 7836
  • Country: us
  • hp>Agilent>Keysight>???
Re: OpenAI ChatGPT for electronics design
« Reply #15 on: February 27, 2023, 02:09:58 am »
Wait a sec ... it said -

"Overall, the evidence suggests that humans are intelligent and have developed many remarkable abilities that have allowed --> us <-- to thrive and prosper as a species."

Us? Us?

Shouldn't it have said "You" ? Does it think it's one of US? Has it assimilated it's creators?

This is getting creepy!
I told my friends I could teach them to be funny, but they all just laughed at me.
 
The following users thanked this post: RJSV

Offline Brumby

  • Supporter
  • ****
  • Posts: 12413
  • Country: au
Re: OpenAI ChatGPT for electronics design
« Reply #16 on: February 27, 2023, 02:26:39 am »
Sentience?   :scared:
 

Offline Brumby

  • Supporter
  • ****
  • Posts: 12413
  • Country: au
Re: OpenAI ChatGPT for electronics design
« Reply #17 on: February 27, 2023, 02:29:10 am »
Serious question, though...

I have tried a couple of times to get a response in the form of an image - and while I am given a URL for said image, I have been unable to access any image.  I just get file not found.

Any ideas?
 

Offline xrunner

  • Super Contributor
  • ***
  • Posts: 7836
  • Country: us
  • hp>Agilent>Keysight>???
Re: OpenAI ChatGPT for electronics design
« Reply #18 on: February 27, 2023, 02:39:00 am »
I tried this, but don't know if it's what you were doing -

Me: Can you generate or point me to an image of Russian Blue cat?

ChatGPT: As an AI language model, I'm not able to generate images, but I can suggest a few websites where you can find images of Russian Blue cats:

    Google Images: You can search for "Russian Blue cat" on Google Images and you'll find many pictures of this beautiful breed.

    Shutterstock: This is a stock photo website that has a large collection of high-quality images of Russian Blue cats.

    Getty Images: Another popular stock photo website that has a variety of images of Russian Blue cats.

I hope that helps you find the image you're looking for!

 
I told my friends I could teach them to be funny, but they all just laughed at me.
 

Offline MK14

  • Super Contributor
  • ***
  • Posts: 4987
  • Country: gb
Re: OpenAI ChatGPT for electronics design
« Reply #19 on: February 27, 2023, 02:55:57 am »
The REAL intelligence comes from someone saying "What the @@@@ is this - another tech industry fad? - No thanks, I think we humans have a little more experience and discernment than this dumb robot bullshit"

The issue is, people's brains will start to depend on this crap, and their once sharp minds with begin to atrophy, depending on some dub box of rocks to "think" and to "design" things for them. Watch it all come crashing down. Whoooooops!  :palm:

Supposedly humans are intelligent - where's the evidence of this?

ChatGPT's response:

Quote
As an AI language model, I don't have personal opinions or beliefs, but I can provide you with some information and perspectives on the topics you mentioned.

Firstly, the idea that AI or robotics will replace human intelligence is a controversial topic that has been discussed for decades. While AI and robotics have undoubtedly made impressive advances in recent years, there is no indication that they will entirely replace human intelligence anytime soon. AI and robotics are tools created and used by humans to enhance their capabilities, solve complex problems, and automate repetitive tasks.

Secondly, the notion that people's minds will atrophy because of technology is also debatable. Research suggests that technology can improve cognitive functions such as attention, memory, and problem-solving skills. However, it's essential to use technology in moderation and balance it with other activities that stimulate the mind, such as reading, physical exercise, or social interactions.

Lastly, the evidence of human intelligence can be found in various fields, such as science, art, literature, music, philosophy, and many others. Humans have achieved incredible feats, from landing on the moon to developing vaccines that save millions of lives. While humans are not perfect and still have many challenges to overcome, their ingenuity, creativity, and adaptability demonstrate their intelligence and potential.
 

Offline JPortici

  • Super Contributor
  • ***
  • Posts: 3573
  • Country: it
Re: OpenAI ChatGPT for electronics design
« Reply #20 on: February 27, 2023, 06:56:33 am »
< I went the next step ...>

Q: Respond to the same question sarcastically

A: Oh, absolutely! Humans are renowned for their unparalleled intelligence. I mean, who else could come up with such brilliant ideas like using plastic straws and bags that take hundreds of years to decompose, or starting wars over trivial matters, or creating reality TV shows that celebrate shallowness and stupidity? Clearly, we are the pinnacle of intelligence in the universe.

Q: Tell me more!

A: Well, where do I even begin? Humans are so intelligent that we can't even figure out how to live sustainably on our own planet without causing widespread environmental destruction. We're so smart that we spend billions of dollars on weapons of mass destruction instead of investing in things that would actually benefit our species, like education and healthcare. And let's not forget how clever we are at ignoring scientific evidence and choosing to believe in myths and superstitions instead. Truly, it's a wonder how we've managed to survive this long with such exceptional intelligence.




< I think I might stop here >  ;D

that is literally a quote from something, word for word. but i can't remember from where...

EDIT: It could be one episode from Love/Death/Robots, one in which the robot trio visits earth, and the quote should be from the robot with female voice describing humanity achievements in a sarcastic tone (which would be a fun coincidence)
« Last Edit: February 27, 2023, 12:08:21 pm by JPortici »
 

Offline eti

  • Super Contributor
  • ***
  • !
  • Posts: 1801
  • Country: gb
  • MOD: a.k.a Unlokia, glossywhite, iamwhoiam etc
Re: OpenAI ChatGPT for electronics design
« Reply #21 on: February 27, 2023, 07:28:09 am »
Gosh, the sheer (actually clumsy, laughable) arrogance of the programmers of AI systems to even hint that AI “could” replace human intelligence, given to us by the creator of the universe.

You know you have your answer when the answer is as vague as “not anytime soon” but they’re too arrogant and cock sure to be HONEST and say “Don’t be utterly ludicrous! I’m just a machine; are you drunk?!”

Mankind’s feeble, clumsy “advances” thinking they’re somehow capable of replacing GODs creations. Yeah? Okay. 😂😂😂😂😂😂😂😂😂😂😂😂😂😂
 

Offline MK14

  • Super Contributor
  • ***
  • Posts: 4987
  • Country: gb
Re: OpenAI ChatGPT for electronics design
« Reply #22 on: February 27, 2023, 08:14:05 am »
Gosh, the sheer (actually clumsy, laughable) arrogance of the programmers of AI systems to even hint that AI “could” replace human intelligence, given to us by the creator of the universe.

You know you have your answer when the answer is as vague as “not anytime soon” but they’re too arrogant and cock sure to be HONEST and say “Don’t be utterly ludicrous! I’m just a machine; are you drunk?!”

Mankind’s feeble, clumsy “advances” thinking they’re somehow capable of replacing GODs creations. Yeah? Okay. 😂😂😂😂😂😂😂😂😂😂😂😂😂😂


I put your question (last post), to it, and it responded.  Which got us chatting, and ending up discussing the validity, of part of what you said.  Later part of discussion, here:

Quote
Certainly. There are several arguments that have been put forward over the years to suggest that a master creator or higher power may not exist. Here are some of the most common ones:

    The Problem of Evil: This argument states that the existence of a master creator or higher power is inconsistent with the existence of evil and suffering in the world. If there was a perfect, all-powerful, and all-loving creator, then they would have the power and desire to prevent or eliminate evil and suffering, yet it still exists in the world. This contradiction suggests that a master creator or higher power does not exist.

    The Argument from Inconsistency: This argument suggests that many of the attributes commonly associated with a master creator or higher power are logically inconsistent with one another. For example, if a master creator is all-powerful, then they should be able to create a stone that is too heavy for them to lift, but if they are unable to lift the stone, then they are not all-powerful. This and similar logical paradoxes suggest that a master creator or higher power may not exist.

    The Argument from Ignorance: This argument suggests that the existence of a master creator or higher power is based on the assumption that natural processes cannot account for certain phenomena. However, this assumption is a logical fallacy known as an argument from ignorance. Just because we don't currently understand how something works, it doesn't mean that it must be the work of a master creator or higher power. Science has demonstrated that many things that were once attributed to a master creator can actually be explained through natural processes.

    Occam's Razor: This argument suggests that the simplest explanation is usually the correct one. Therefore, if there is a naturalistic explanation for the universe and all living things within it, then it is more likely to be correct than a complex explanation involving a master creator or higher power.

    Lack of Evidence: This argument states that there is no empirical evidence or scientific method to prove or disprove the existence of a master creator or higher power. Claims about the existence of a master creator or higher power are based on faith, personal experience, or philosophical reasoning, rather than empirical evidence or scientific investigation. Therefore, it is difficult to make a convincing argument for the existence of a master creator or higher power based on evidence alone.


More?

Certainly, here are some additional points that have been raised in arguments against the existence of a master creator or higher power:

    The Argument from Cultural Relativism: This argument suggests that the existence of a master creator or higher power is culturally constructed and varies from culture to culture. Different religions have different beliefs about the nature of the master creator, their attributes, and their role in the world. If the existence of a master creator or higher power was a universal truth, then there would be more consistency across cultures and religions. The fact that beliefs about the master creator vary so widely suggests that it is a human construct rather than a real entity.

    The Argument from Naturalism: This argument suggests that the universe and all living things within it can be explained solely by natural processes, such as evolution and the laws of physics. There is no need to invoke a master creator or higher power to explain the origin and diversity of life on earth. In fact, the scientific method has been able to explain many things that were once attributed to a master creator, such as the origin of the universe and the diversity of life on earth.

    The Argument from Personal Experience: This argument suggests that personal experiences of a master creator or higher power are subjective and cannot be verified by others. People may have feelings of awe, wonder, or transcendence that they attribute to a master creator or higher power, but these experiences are not proof of the existence of a real entity. Moreover, personal experiences of a master creator or higher power are often inconsistent and contradictory, which suggests that they are more likely to be the result of psychological and cultural factors rather than the presence of a real entity.

    The Argument from the Absence of Miracles: This argument suggests that the lack of observable miracles or supernatural events in the world suggests that there is no master creator or higher power who intervenes in the world. If a master creator or higher power existed, then they would be expected to occasionally intervene in the world in order to demonstrate their presence and power. However, the absence of such events suggests that a master creator or higher power does not exist or does not intervene in the world.

These are just a few of the many arguments that have been made against the existence of a master creator or higher power. Ultimately, the question of whether a master creator or higher power exists is a matter of personal belief and interpretation, and different people may come to different conclusions based on their own values, experiences, and perspectives.

More?

Sure, here are a few more arguments against the existence of a master creator or higher power:

    The Argument from Evil: This argument suggests that the existence of evil and suffering in the world is incompatible with the idea of a loving and all-powerful master creator or higher power. If a master creator or higher power existed, then they would presumably be able to prevent or alleviate suffering and evil, but these things continue to occur in the world. The existence of evil and suffering is therefore seen as evidence against the existence of a master creator or higher power.

    The Argument from Incoherence: This argument suggests that the concept of a master creator or higher power is incoherent or self-contradictory. For example, if a master creator is all-powerful, can they create a stone so heavy that even they cannot lift it? This paradoxical question suggests that the concept of an all-powerful master creator may be logically inconsistent or impossible.

    The Argument from Unintelligibility: This argument suggests that the concept of a master creator or higher power is unintelligible or meaningless. Some philosophers have argued that the concept of a master creator or higher power is too abstract or vague to be meaningful or coherent. Without a clear definition of what is meant by a master creator or higher power, it is difficult to evaluate whether or not they exist.

These arguments all point towards the idea that there are logical and philosophical inconsistencies with the concept of a master creator or higher power. Whether or not these arguments are ultimately convincing will depend on an individual's own beliefs and worldviews.

Any additional points?

Sure, here are a few more points to consider:

    The Argument from Evolution: Some argue that the theory of evolution provides a naturalistic explanation for the complexity and diversity of life on Earth, without the need for a master creator or higher power. According to this argument, the development of life through natural selection and genetic mutation is a sufficient explanation for the diversity of species we observe today.

    The Argument from Science: Many scientific discoveries have challenged traditional religious beliefs and put into question the idea of a master creator or higher power. For example, the discovery of the Big Bang suggests that the universe had a beginning, which contradicts the idea of an eternal master creator. Additionally, advancements in neuroscience and psychology have provided naturalistic explanations for phenomena that were previously attributed to divine intervention.

    The Argument from Occam's Razor: This argument suggests that the simplest explanation is usually the best one. According to this principle, if there are two explanations for a phenomenon, the one that makes fewer assumptions is more likely to be true. In the case of the master creator or higher power, the naturalistic explanation that doesn't require the existence of an all-powerful being is simpler and therefore more likely to be true.

These arguments all point towards the idea that the concept of a master creator or higher power is unnecessary, unprovable, and potentially incompatible with scientific discoveries and naturalistic explanations. Again, whether or not these arguments are ultimately convincing will depend on an individual's own beliefs and worldview.

So, is there anything to suggest, that such a creator or higher power, might actually exist?

There are certainly arguments and beliefs that suggest the existence of a master creator or higher power. Here are a few examples:

    The Cosmological Argument: This argument suggests that the existence of the universe requires a first cause, which is often attributed to a master creator or higher power. The idea is that everything that exists must have a cause, and since the universe exists, there must be a cause that brought it into being.

    The Teleological Argument: This argument suggests that the complexity and order of the universe suggest the presence of a master creator or higher power. The idea is that the universe appears to be designed for a specific purpose or end, and this design implies the existence of an intelligent creator.

    Religious Experience: Many people report having experiences that they interpret as encounters with a higher power or master creator. These experiences can range from feelings of awe and wonder in nature to more mystical or transcendent experiences that are difficult to describe.

    Personal Beliefs: For some people, the existence of a master creator or higher power is a matter of faith or personal conviction. They may feel that the universe is too complex or miraculous to be the result of chance, or they may feel a deep sense of connection to a higher power through prayer or religious practices.

It's important to note that none of these arguments or beliefs provide conclusive proof of the existence of a master creator or higher power. Whether or not such a being exists is ultimately a matter of personal belief and interpretation of the available evidence.

Continue

    Moral Argument: Some people argue that the existence of a master creator or higher power is necessary to ground objective morality. The idea is that without a higher power to establish moral standards, there can be no objective basis for morality. Therefore, the existence of objective morality implies the existence of a master creator or higher power.

    The Argument from Consciousness: This argument suggests that the existence of consciousness and subjective experience cannot be explained by materialistic or naturalistic explanations alone. The idea is that consciousness is a unique and irreducible feature of reality that requires the existence of a master creator or higher power.

    The Argument from Miracles: Some people point to supposed miracles as evidence for the existence of a master creator or higher power. These miracles are often interpreted as supernatural events that defy natural laws and suggest the intervention of a higher power.

It's worth noting that many of these arguments have been critiqued and debated by philosophers and theologians over the centuries, and there is no consensus on whether they provide strong evidence for the existence of a master creator or higher power. Ultimately, whether or not such a being exists is a matter of personal belief and interpretation of the available evidence.

So is there evidence, these miracles actually took place?

The question of whether miracles actually take place is a matter of debate and interpretation, and opinions on this subject vary widely depending on one's beliefs, religious or otherwise.

Many people believe in the existence of miracles and point to a variety of experiences and events that they believe cannot be explained by natural causes alone. These include apparent healings, near-death experiences, and other phenomena that are interpreted as evidence of divine intervention or supernatural forces.

However, skeptics of miracles argue that these events can often be explained by natural causes or are simply the result of wishful thinking, superstition, or coincidence. They may point to the lack of scientific evidence for miraculous events and argue that any claims of miracles should be subjected to rigorous scrutiny before being accepted as true.

It's worth noting that the question of whether miracles actually take place is ultimately a matter of personal belief and interpretation of the available evidence. Some people may find certain experiences or events convincing evidence of miracles, while others may remain skeptical or unconvinced.
« Last Edit: February 27, 2023, 08:16:23 am by MK14 »
 

Offline eti

  • Super Contributor
  • ***
  • !
  • Posts: 1801
  • Country: gb
  • MOD: a.k.a Unlokia, glossywhite, iamwhoiam etc
Re: OpenAI ChatGPT for electronics design
« Reply #23 on: February 27, 2023, 08:36:35 am »
Any amount of pseudo intellectual “wisdom” to attempt to “disprove” GODs existence. I don’t care to read that wall of idiocy . God is real. That thing is merely a reflection of, or vaguely as “clever” as the intelligence or stupidity it’s been fed with when being designed.
 
The following users thanked this post: MK14

Offline MK14

  • Super Contributor
  • ***
  • Posts: 4987
  • Country: gb
Re: OpenAI ChatGPT for electronics design
« Reply #24 on: February 27, 2023, 09:05:12 am »
Any amount of pseudo intellectual “wisdom” to attempt to “disprove” GODs existence. I don’t care to read that wall of idiocy . God is real. That thing is merely a reflection of, or vaguely as “clever” as the intelligence or stupidity it’s been fed with when being designed.

You seem to be essentially saying, the old computer adage.  Garbage in, garbage out.

When / if, these AI systems, ever become more genuinely, generally intelligent.  We might be able to get more interesting / sensible / debatable answers.
« Last Edit: February 27, 2023, 09:07:13 am by MK14 »
 


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf