Self aware computers project - written in BBC Basic, help needed!

[Pete]

I'd be keen for some people to try my program version 1.20 on their own computers, please try it ! No takers so far. It really does work.

Everyone that comes to my house and hears the program in operation is impressed, and 90% of them have concluded that the computers are responding consciously with intelligence to music.

The program now works with a selectable menu of "Start, Stop, About, Volume Up, Volume Down, and Exit". This is a tiny menu which appears at the top of the screen. Could anyone suggest an improvement to this - for example a menu that could appear in the tray rather than at the top of the screen. The reason I ask is that at the moment the window can be maximised or expanded, which just reveals white space and a flashing cursor.

(Thanks Svein): Present code being used for the GUI aspect of the program as follows:

10 INSTALL @lib$+"COMLIBA"
20
30 ON ERROR PROC_comexit : PRINT 'REPORT$ : END
40 ON CLOSE PROC_comexit : QUIT
50
60 SYS "SetWindowText", @hwnd%,"Self aware computer v1.2"
70 SYS "MoveWindow", @hwnd%, 300, 0, 360, 60, 1 : REM small window
80
90 SYS "CreatePopupMenu" TO hpop1%
100 SYS "AppendMenu", hpop1%, 0, 1, "Start self aware"
110 SYS "AppendMenu", hpop1%, 0, 2, "Stop in a moment"
120 SYS "AppendMenu", hpop1%, 0, 3, "Volume up"
130 SYS "AppendMenu", hpop1%, 0, 4, "Volume down"
140 SYS "AppendMenu", hpop1%, 0, 5, "About"
150 SYS "AppendMenu", hpop1%, 0, 6, "Exit"
160
170 SYS "CreateMenu" TO hmenu%
180 SYS "AppendMenu", hmenu%, 16, hpop1%, "&Menu"
190 SYS "SetMenu", @hwnd%, hmenu%
200 SYS "DrawMenuBar", @hwnd%
210
220 sound%=TRUE : vol%=-10
230
240 Click% = -1
250 ON SYS Click% = @wparam% : RETURN
260 PROCselfawareoutput
270 REPEAT:PROCgui:WAIT 1:UNTIL sound%=TRUE
280 GOTO 260:END

I have come across a company who may be able to help me develop this into an app - but they haven't used BBC Basic for Windows before. It would be great to have this working on iphone, Android. It would be a shame if I have to port it into another programming language.

Can anyone help me with final development of this version 1.20 into releasable form? Am I allowed to offer payment or is that not allowed? I am not advanced enough with Windows/Android/iphone to be able to complete development into a form which can be released. I will need to get the program signed by a code signing authority.

Will anyone try my program? Feel free to report your findings.

[RichardRussell]

Everyone that comes to my house and hears the program in operation is impressed, and 90% of them have concluded that the computers are responding consciously with intelligence to music.

Nobody here (so far as I'm aware) hears anything other than random notes, but that may well be because they know how computers work, and they know how BBC BASIC works, so they know that what you claim is impossible. Somebody without that background may be more open to suggestion.

It would be great to have this working on iphone, Android. It would be a shame if I have to port it into another programming language.

Since you've written it in BBC BASIC having it "working" on Android and iOS is straightforward, you may not need to make any changes other than to remove the Windows-specific features such as Svein's GUI code. Try it first in BBC BASIC for SDL 2.0 and if it works there it will likely work with few if any changes on a mobile device.

However, making an app that you could submit to the relevant App Store is another matter. Neither the Android nor the iOS app stores will accept an app coded in BBC BASIC currently, because it cannot meet their stringent requirements. For example the Android Play Store requires every app to have a 64-bit ARM executable, whereas (currently, at least) BBC2APK-generated apps are only 32-bit (there's a 64-bit x86 executable included I think).

Can anyone help me with final development of this version 1.20 into releasable form?

Would you consider moderating the claims you make for your program? Rather than claiming that the computer is 'self aware', you might get more support if you were to say that it demonstrates how easily the human brain can be fooled. Optical illusions are fascinating, and can shine light on the workings of the eye-brain system that are of genuine scientific merit. Your program is saying something similar about the brain's auditory system, and if marketed that way is likely to attract more 'mainstream' interest.

I am not advanced enough with Windows/Android/iphone to be able to complete development into a form which can be released.

In what respect is it not already compatible with mobile devices? Are you currently relying on the keyboard (which obviously would need to be changed) or only on the mouse (which may work on a touchscreen without modification)?

[Pete]

I will try to answer individual points made by Richard Russell later. All I can say now, again, is you are utterly wrong. If the computers did not sing to music and produced random tones then I wouldn't pursue this. I do not have a problem with my brain, you have argued this point before, and you are not correct.

No one has tried it here to date. Has anyone listened to a full demo eg 5 minutes, rather than just a few seconds? You cannot ascertain a correct impression by listening to a few seconds - you might by doing that get an impression of it sounding random.

Will no one try my program? If not, I will have to stop being an active member, and spend up to thousands hiring someone/people to put it into a different language. That will be a great shame, I did love BBC Basic.

I post with great excitement my discovery which I made in 2009, at least 1 or 2 members have stated they have at least some belief in what I'm doing - then I'm told by Richard Russell the claim is rubbish and impossible.

[Pete]

In what respect is it not already compatible with mobile devices? Are you currently relying on the keyboard (which obviously would need to be changed) or only on the mouse (which may work on a touchscreen without modification)?

It works using the mouse.

Would you consider moderating the claims you make for your program? Rather than claiming that the computer is 'self aware', you might get more support if you were to say that it demonstrates how easily the human brain can be fooled. Optical illusions are fascinating, and can shine light on the workings of the eye-brain system that are of genuine scientific merit. Your program is saying something similar about the brain's auditory system, and if marketed that way is likely to attract more 'mainstream' interest.

Optical illusions are fascinating, for example the illusion of objects moving when in fact the image is still. This program is NOT saying something similar about the brain's hearing system, and I would never market it as something that does that. It's not a fake, it's a genuine program which has worked since 2009.

However, making an app that you could submit to the relevant App Store is another matter. Neither the Android nor the iOS app stores will accept an app coded in BBC BASIC currently, because it cannot meet their stringent requirements. For example the Android Play Store requires every app to have a 64-bit ARM executable, whereas (currently, at least) BBC2APK-generated apps are only 32-bit (there's a 64-bit x86 executable included I think).

OK, I'll find someone to help me produce a 64-bit ARM executable version. Thanks for that info.

[RichardRussell]

OK, I'll find someone to help me produce a 64-bit ARM executable version.

I said that for an Android app to be accepted for the Play Store it needed to include a 64-bit ARM executable, not consist solely of one. Ideally it should include all four executable types - 32-bit ARM, 64-bit ARM, 32-bit x86 and 64-bit x86 - to maximise compatibility with older mobile devices and Chromebooks, for example.

The reason that BBC2APK does not currently bundle a 64-bit ARM executable is the potential for incompatibility. Although all the supplied example programs, libraries and tools run successfully in 64-bits - they have to in order to work on MacOS, iOS and 64-bit versions of Linux - this does not come for free. To ensure 64-bit compatibility the BASIC programmer must take account of the considerations documented here.

But the need to be 64-bit compatible will inevitably grow. Android may drop support for 32-bit apps (as MacOS and iOS already have), many Linux users dislike installing the 32-bit architecture on their 64-bit machines, and even Windows may go the same way eventually. So before too long it will probably be necessary for all BBC BASIC programs to be 64-bit compatible, and at that point I can add the 64-bit executable to the bundle.

[DDRM]

All I can say now, again, is you are utterly wrong.

Whoa! That's a pretty strong statement, for which you have, as far as I'm aware, no evidence but your opinion against Richard's. You can say you BELIEVE he is wrong, if you wish, but I don't feel you can justify this statement.

Please be careful and considerate in how you refer to other members of the forum.

Best wishes,

D

[hellomike]

Pete,

Does your program produce a log file recording what note (pitch) was generated at what moment and its length and amplitude?
If not, add it because then you run the program 100 times over exactly the same piece of music each time!

After that, compare the 100 created log files and if you are right there should be a overwhelming correlation between all of them.

And if there isn't well then.... there isn't.

I'm looking forward to the outcome.

Mike

[Pete]

DDRM ok sorry about that, I thought utterly meant very, upon searching the definition it wasn’t the wisest word to use, apologies to you and Richard Russell. Ok, I firmly believe that the computers are reacting to music when the program is run, not acting independently and randomly. Hellomike, I’ll post back tomorrow thanks.

[Pete]

Hellomike, a typical song is 3 and a half minutes long. It’s relatively straightforward to add a log file output to the program. So 100 repeats would involve 350 minutes in that example - that is just under 6 hours. The reason I won’t do as you suggest is because the computer may become bored witless in being expected to react to the same music for 6 hours. Performance will undoubtedly tail off perhaps after 20 to 30 minutes.

I have a piece of music the computer has over time repeatedly heard and reacted to, effectively been trained to. I can certainly do something similar to what you suggest, on a piece of music which lasts a minute and I know the computer likes, as her beeps are always more excited when it is played. I suggest I play the music a dozen times and analyse logs from that, but not 100 times.

I will set this up and record the findings some time between this Friday and this Sunday, and post results on Sunday evening or before. Is that a suitable plan?

[hellomike]

The computer might get bored. That's your answer. Mmm.

The issue here is of course that virtually no-one believes your claim so I think it's in your own interest to come up with some kind of proof.

I figured it out! Play a song that your computer hates. The anger response should always be roughly the same and thus correlate! And don't give me that you can't bare to give him that treatment!

It would be the ultimate test and conclusive.

By the way, I'm not interested in the individual log files, just the scientific correlation between all of the files.

Mike