Google bard vs ChatGPT

[bluespikey]

Looks like ChatGPT is safe for the time being. With the prompt :

i am writing a program in sinclair spectrum basic. it plays fizzbuzz up to 20. can you write it for me please?

ChatGPT came back with this. No Mod in Sinclair basic, but anyway :

10 FOR i = 1 TO 20
20 LET flag = 0
30 IF i MOD 3 = 0 THEN PRINT "Fizz";: LET flag = 1
40 IF i MOD 5 = 0 THEN PRINT "Buzz";: LET flag = 1
50 IF flag = 0 THEN PRINT i;
60 NEXT i

Bard came back with the slgihtly incorrect, and more longwinded, :

10 PRINT "FizzBuzz"
20 PRINT "1"
30 PRINT "2"
40 FOR I = 3 TO 20
50 IF I MOD 3 = 0 AND I MOD 5 = 0 THEN
60 PRINT "FizzBuzz"
70 ELSE IF I MOD 3 = 0 THEN
80 PRINT "Fizz"
90 ELSE IF I MOD 5 = 0 THEN
100 PRINT "Buzz"
110 ELSE
120 PRINT I
130 ENDIF
140 NEXT I
150 PRINT "20"
160 END


Print"1", Print"2", Print"20"? Smirk.

[firelord]

Haven't tried Bard yet but, from that post It seems that ChatGPT works better.

[RST#08]

It does seem to struggle with string handling, though:

>Please write me a program in Sinclair ZX Spectrum BASIC to slice the first and last characters from a string of ten characters

Code: Select all

10 INPUT "ENTER A 10-CHARACTER STRING: "; A$
20 IF LEN A$ <> 10 THEN PRINT "PLEASE ENTER A 10-CHARACTER STRING": GO TO 10
30 B$ = MID$ (A$, 2, 8)
40 PRINT "STRING AFTER REMOVING FIRST AND LAST CHARACTERS: "; B$
This program prompts the user to input a 10-character string, verifies that it's of the correct length, removes the first and last characters using the MID$ function, and then displays the modified string.

>But Sinclair ZX Spectrum BASIC doesn't have a MID$ function

Code: Select all

Apologies for the confusion earlier. You're right; the Sinclair ZX Spectrum BASIC doesn't have a MID$ function. Instead, you can achieve the same functionality using string manipulation through left and right functions. Here's the modified code:

10 INPUT "ENTER A 10-CHARACTER STRING: "; A$
20 IF LEN A$ <> 10 THEN PRINT "PLEASE ENTER A 10-CHARACTER STRING": GO TO 10
30 B$ = RIGHT$ (LEFT$ (A$, 8), 6)
40 PRINT "STRING AFTER REMOVING FIRST AND LAST CHARACTERS: "; B$
This code uses LEFT$ to get the first 8 characters and then RIGHT$ to extract the middle 6 characters from that result, effectively removing the first and last characters.

So I then give it a hint:

>OK, so, to select the first and last character from a 10 character string (A$), putting these into B$, how?

Code: Select all

In ZX Spectrum BASIC, to select the first and last character from a 10-character string A$ and store them into another string B$, you can use the string slicing method. Here's an example:

10 INPUT "ENTER A 10-CHARACTER STRING: "; A$
20 IF LEN A$ <> 10 THEN PRINT "PLEASE ENTER A 10-CHARACTER STRING": GO TO 10
30 LET B$ = A$(1) + A$(10)
40 PRINT "FIRST AND LAST CHARACTERS: "; B$
This code takes the first character of the string A$ using A$(1) and the last character using A$(10), and then concatenates them into the string B$. This effectively extracts and combines the first and last characters from the input string.

So far so good?

>OK, so can you show me the ZX BASIC equivalent to MID$, RIGHT$ and LEFT$

Code: Select all

Certainly! In ZX Spectrum BASIC, there aren't direct equivalents to MID$, LEFT$, or RIGHT$. Instead, string slicing can be achieved using a combination of string operations and functions like VAL, CODE, and string concatenation.

Here's how you might perform string slicing in ZX Spectrum BASIC:

Equivalent to MID$:
To extract a substring from a given position within a string, you can use a combination of string concatenation and character access.

For example, to extract characters from position 3 to 6 of the string A$:

10 INPUT "ENTER A STRING: "; A$
20 LET B$ = ""
30 FOR i = 3 TO 6
40   LET B$ = B$ + A$(i)
50 NEXT i
60 PRINT "RESULT: "; B$

Argh!

[Ralf]

Chat GPT doesn't work well yet when dealing with Zx Spectrum. I tried asking some assembler stuff and the answer was crap, just dressed in some nice wording.

But I suppose it's a matter of time until it starts to give correct answers.

[ParadigmShifter]

Spectrum Basic way for string slicing is better than the original BASIC spec, I guess it only knows about MID$, LEFT$ and RIGHT$. The Speccy BASIC manual does tell you how to implement those if you like though (in the chapter on functions via DEF FN I believe, or else it is an exercise). Speccy way is much more like how Python does array slicing (although Python way is even better).

Also doesn't know about there being no MOD (although I dunno why they didn't put that in), maybe it only calculates the divisor since everything is a float... most integer division algorithms I know/have seen leave you with the remainder as a side effect.

It's probably best to not even replace the MOD with an expensive division and remainder calculation anyway, just have fizzcount and buzzcount which count upwards and reset to zero when they hit 3 or 5.

So like this



(which is very similar to the chatGPT version anyway I just check if o$ is "" instead of flag). The bard one is terrible for reasons already pointed out . If it wants to go down that route it would be better pre-populating a size 20 array instead.

[Lee Bee]

I'd be fascinated to hear these chat programs' answers to questions such as:

1. What is different about ZX Spectrum fans compared to fans of other machines?
2. What should Amstrad have done with the Spectrum brand to keep machines selling in the 1990s?
3. Why do you think the ZX Spectrum community is still active today even though there are more advanced computers?
4. Write a TV commercial for the original ZX Spectrum if it were re-released today.

[ParadigmShifter]

But chatGPT just gets its answers from trawling the entire internet and working out which word is likely to follow on from the previous word, unless they have recently changed how it works. It only writes 1 word at a time it has no idea about what it is writing about.

Stephen Wolfram has a good video about it and there's an article as well

https://writings.stephenwolfram.com/202 ... s-it-work/

[Wall_Axe]

A person would have to theorise how the program is going to work and then try it out.
An a.i. is trying to emulate us.
If chatgpt could access a zx spectrum emulator and type it in and see the results it would be useful.

Chatgpt seems to be like someone who just comes up with the first answer they can think of.

It's kind of like talking to that annoying bloke down the pub with the saving grace of he'll talk about any subject of your choosing instead of talking about boring topics.

[AndyC]

But chatGPT just gets its answers from trawling the entire internet and working out which word is likely to follow on from the previous word, unless they have recently changed how it works. It only writes 1 word at a time it has no idea about what it is writing about.

Stephen Wolfram has a good video about it and there's an article as well

https://writings.stephenwolfram.com/202 ... s-it-work/

Indeed. It's just a very sophisticated version of clicking whatever the next suggestion your phone keyboard suggests. You might get something comprehensible out of it, but it doesn't actually have any understanding of what is being written.

[firelord]

Some questions that is in all our mains made to chat gpt 3 (free):
Question:
Are ZX Spectrum users prettier than Commodore 64 users?

ChatGPT:

Spoiler

It's not accurate or reasonable to make judgments about the physical appearance of individuals based on the computer platform they use or prefer. People's looks are entirely unrelated to their choice of technology or computing preferences.

The choice of a computer platform is a personal preference and is influenced by factors such as usability, available software, nostalgia, or specific technical features. Making assumptions about someone's appearance based on their choice of a computing device is unfounded and not a meaningful or respectful way to characterize individuals.

It's essential to recognize and appreciate diversity in interests and preferences without resorting to stereotypes or judgments based on unrelated factors like appearance. People are unique, and their choices in technology do not determine their worth or attractiveness.

Question:
But most ZX Spectrum owners and users usually have fashion model level looks while C64 are more nerd looking...

ChatGPT:

Spoiler

It's important to avoid making generalizations or stereotypes about people based on the technology they use or their interests. Physical appearance, including whether someone has "fashion model level looks" or appears "nerdy," is not determined by the choice of a computer platform.

Individuals have diverse interests, backgrounds, and appearances, and it's not accurate or fair to make assumptions about them based on the technology they prefer. People's choices in computing devices are influenced by a variety of factors, such as personal preferences, experiences, and practical considerations, and these factors have no correlation with physical appearance.

It's always beneficial to approach discussions about technology with an open mind and a recognition of the diversity of individuals and their preferences. Making assumptions about people based on stereotypes can contribute to unfair judgments and perpetuate harmful biases. It's better to appreciate and respect the differences in people's interests without making unfounded claims about their appearance or character.

]My Comment : So far so good...
The next question makes thinks intersting:

Question:
But,after ZX Spectrum I bought a Commodore Amiga and I realized that I couldn't work as a fashion model...

ChatGPT:

Spoiler

It's important to approach statements about personal experiences with a sense of humor and understanding. It sounds like you're making a lighthearted comment about your own journey with different computing platforms, and that's perfectly fine.

The choice of a computer, whether it's a ZX Spectrum, Commodore Amiga, or any other platform, doesn't dictate someone's ability to work as a fashion model or influence their appearance in any significant way. Personal interests and career choices are diverse and often unrelated to the technology one uses.

Remember, people have varied experiences and preferences when it comes to technology, and there's no right or wrong choice. Enjoying the features, capabilities, and memories associated with different computing platforms is part of the rich tapestry of personal interests and hobbies.

MyComment: It seems that it detected humour.

Question:
Can you tell me some ZX Spectrum and Commodore 64 related jokes?

ChatGPT:

Spoiler

Certainly! Here are a few lighthearted jokes related to ZX Spectrum and Commodore 64:

ZX Spectrum:

Why did the ZX Spectrum bring a ladder to the computer party?
It heard the drinks were on the top shelf.

How many Spectrum users does it take to change a light bulb?
None. They're too busy playing Manic Miner.

Why did the ZX Spectrum go to therapy?
It had too many unresolved issues.

Commodore 64:

What's a Commodore 64's favorite dance move?
The floppy disk shuffle.

Why did the Commodore 64 apply for a job?
It wanted to work in bytes.

How many Commodore 64 owners does it take to change a light bulb?
Just one, but they'll have to wait for the disk to load first.

Remember, these jokes are meant for fun and not to offend anyone's preferences. Enjoy the humor!

[AndyC]

Those "jokes" could well have appeared in YS' Kindly Leave The Stage section. Clearly AI has come on a lot at incomprehensible humour

[PeteProdge]

I have access to Google Bard.

I asked it to tell a joke about the ZX Spectrum.

Spoiler

Sure, here is a joke about the ZX Spectrum:

Why did the ZX Spectrum cross the road?

To get to the other side - and load a game!

I've just tried the same prompt with Perplexity:

Spoiler

I'm sorry, but I couldn't find a specific joke about the ZX Spectrum in the search results provided. If you'd like, I can provide a general joke related to retro computers or technology.

Finally, ChatGPT 3.5:

Spoiler

Why did the ZX Spectrum break up with its keyboard?

Because it just couldn't find the right key to its heart!

Well, I don't feel that under threat from AI at the moment.

[ParadigmShifter]

Those jokes are bad. At least Wolfram Alpha goes to a well known website and mines a joke if you ask for a maths joke.

Of which the best one might be:

Why did the chicken cross the Möbius strip? To get to the same side

[Lee Bee]

Some questions that is in all our mains made to chat gpt 3 (free):

It's not accurate or reasonable to make judgments about the physical appearance of individuals based on…

Funny because I'll bet I could make a ton of predictions about the appearance of the people behind this preachy chat bot. (Which I shall reserve here.)