[8pla.net]

Mr.SecretHeavenCode

I thought your original question was fine.  If you are a beginner, there is no shame in that. And I mention this without implying irony or mockery to anyone on this thread.

In simple terms A.I. is made with input to a machine, which is called a "stimulus" that the machine matches on to print an output called a "response".  The next stage is to put this process through a loop to simulate an ongoing conversation back and forth.

Pseudocode:  Handles HELLO, GOODBYE and can TEST fifty other unique stimuli.

Code

if STIMULUS = "HELLO" then RESPONSE = "Hi There!"  
if STIMULUS = 'TEST 1' then RESPONSE = 'Test #1 complete.'
if STIMULUS = 'TEST 2' then RESPONSE = 'Test #2 complete.'
if STIMULUS = 'TEST 3' then RESPONSE = 'Test #3 complete.'
if STIMULUS = 'TEST 4' then RESPONSE = 'Test #4 complete.'
if STIMULUS = 'TEST 5' then RESPONSE = 'Test #5 complete.'
if STIMULUS = 'TEST 6' then RESPONSE = 'Test #6 complete.'
if STIMULUS = 'TEST 7' then RESPONSE = 'Test #7 complete.'
if STIMULUS = 'TEST 8' then RESPONSE = 'Test #8 complete.'
if STIMULUS = 'TEST 9' then RESPONSE = 'Test #9 complete.'
if STIMULUS = 'TEST 10' then RESPONSE = 'Test #10 complete.'
if STIMULUS = 'TEST 11' then RESPONSE = 'Test #11 complete.'
if STIMULUS = 'TEST 12' then RESPONSE = 'Test #12 complete.'
if STIMULUS = 'TEST 13' then RESPONSE = 'Test #13 complete.'
if STIMULUS = 'TEST 14' then RESPONSE = 'Test #14 complete.'
if STIMULUS = 'TEST 15' then RESPONSE = 'Test #15 complete.'
if STIMULUS = 'TEST 16' then RESPONSE = 'Test #16 complete.'
if STIMULUS = 'TEST 17' then RESPONSE = 'Test #17 complete.'
if STIMULUS = 'TEST 18' then RESPONSE = 'Test #18 complete.'
if STIMULUS = 'TEST 19' then RESPONSE = 'Test #19 complete.'
if STIMULUS = 'TEST 20' then RESPONSE = 'Test #20 complete.'
if STIMULUS = 'TEST 21' then RESPONSE = 'Test #21 complete.'
if STIMULUS = 'TEST 22' then RESPONSE = 'Test #22 complete.'
if STIMULUS = 'TEST 23' then RESPONSE = 'Test #23 complete.'
if STIMULUS = 'TEST 24' then RESPONSE = 'Test #24 complete.'
if STIMULUS = 'TEST 25' then RESPONSE = 'Test #25 complete.'
if STIMULUS = 'TEST 26' then RESPONSE = 'Test #26 complete.'
if STIMULUS = 'TEST 27' then RESPONSE = 'Test #27 complete.'
if STIMULUS = 'TEST 28' then RESPONSE = 'Test #28 complete.'
if STIMULUS = 'TEST 29' then RESPONSE = 'Test #29 complete.'
if STIMULUS = 'TEST 30' then RESPONSE = 'Test #30 complete.'
if STIMULUS = 'TEST 31' then RESPONSE = 'Test #31 complete.'
if STIMULUS = 'TEST 32' then RESPONSE = 'Test #32 complete.'
if STIMULUS = 'TEST 33' then RESPONSE = 'Test #33 complete.'
if STIMULUS = 'TEST 34' then RESPONSE = 'Test #34 complete.'
if STIMULUS = 'TEST 35' then RESPONSE = 'Test #35 complete.'
if STIMULUS = 'TEST 36' then RESPONSE = 'Test #36 complete.'
if STIMULUS = 'TEST 37' then RESPONSE = 'Test #37 complete.'
if STIMULUS = 'TEST 38' then RESPONSE = 'Test #38 complete.'
if STIMULUS = 'TEST 39' then RESPONSE = 'Test #39 complete.'
if STIMULUS = 'TEST 40' then RESPONSE = 'Test #40 complete.'
if STIMULUS = 'TEST 41' then RESPONSE = 'Test #41 complete.'
if STIMULUS = 'TEST 42' then RESPONSE = 'Test #42 complete.'
if STIMULUS = 'TEST 43' then RESPONSE = 'Test #43 complete.'
if STIMULUS = 'TEST 44' then RESPONSE = 'Test #44 complete.'
if STIMULUS = 'TEST 45' then RESPONSE = 'Test #45 complete.'
if STIMULUS = 'TEST 46' then RESPONSE = 'Test #46 complete.'
if STIMULUS = 'TEST 47' then RESPONSE = 'Test #47 complete.'
if STIMULUS = 'TEST 48' then RESPONSE = 'Test #48 complete.'
if STIMULUS = 'TEST 49' then RESPONSE = 'Test #49 complete.'
if STIMULUS = 'TEST 50' then RESPONSE = 'Test #50 complete.'
if STIMULUS = "GOOD BYE" then RESPONSE = "Please come back soon." 

print RESPONSE

This is Pseudocode which is intended as a simple explanation, and is not at all meant to run as a computer program in this form.  So, please don't bother suggesting any syntax errors.  I am already aware,  but I am going for ease and readability in order to appeal to complete chatbot beginners.

[LOCKSUIT]

No.

And your AI you just said is still nothing 8pla.net.

Man my personality is so fu up here lol. I sound like Dr.Dracula. Sorry I act all ways. You'll never guess anything right.

[LOCKSUIT]

Read above.

I guess I can't tell you's...........I shouldn't.

[ivan.moony]

Just to make sure...

So you're not in some weird mental state, I hope not? Tell me you're ok... cause I had an episode on this subject a few years ago.

[8pla.net]

I guess I can't tell you's...........I shouldn't.

In the popular TV series ‘Humans’, star William Hurt,
the character he plays says there is no 'should' or 'would' in A.I.

Besides, keeping A.I. simple is a challenge in itself.  Complex A.I. has its disadvantages.
According to 'Humans' science fiction, sentience is 17,000 pages of code.
My explanation is less than one page of pseudocode.

[Korrelan]

I think 8pla.net's approach to AGI is valid, in fact all chatbot designs are valid.

In theory any kind of intelligence can be covered by a simple lookup list.  If the list is big enough to cover the domain, and all possible variables are covered. then any combination of inputs can produce a correct intelligent output. The list would need variables/ columns to represent the internal subconscious workings of a mind/ body but the design is still valid.

Even human level intelligence could be covered by a lookup list (albeit huge lol), experience would simply add to the list for the next occurrence of the experience.

If you break neural nets etc, down into their basic parts and functions they are just lookup list routers.  This combination of inputs gives this output. Even deep learning algorithms are just breaking the data down into manageable sections, checking against a lookup list (using a formula) and sending the output to the next lookup list.

Any algorithm/ formula can be represented as a lookup list... think of engineers log table books etc.

You wanted simple; a list is simple.

The hard part is building the list

[8pla.net]

Keeping it simple may help express A.I. in propositional logic by representing pseudocode in mathematical notation:  if P then Q else if R then S else T end if

("HELLO" → "Hi there!") &
(¬"HELLO" & "TEST" → "Test complete.") &
(¬"HELLO" & ¬"TEST" → "Goodbye.")

(P → Q) & (¬P & R → S) & (¬P & ¬R → T)

My improved pseudocode expressed in propositional logic:

(φ→ψ1) ∧(¬φ∧ψ2→ψ3) ∧(¬φ∧ψ3→ψ4)
∧(¬φ∧ψ4→ψ5) ∧(¬φ∧ψ5→ψ6) ∧(¬φ∧ψ6→ψ7)
∧(¬φ∧ψ7→ψ8) ∧(¬φ∧ψ8→ψ9) ∧(¬φ∧ψ9→ψ10)
∧(¬φ∧ψ10→ψ11) ∧(¬φ∧ψ11→ψ12) ∧(¬φ∧ψ12→ψ13)
∧(¬φ∧ψ13→ψ14) ∧(¬φ∧ψ14→ψ15) ∧(¬φ∧ψ15→ψ16)
∧(¬φ∧ψ16→ψ17) ∧(¬φ∧ψ17→ψ18) ∧(¬φ∧ψ18→ψ19)
∧(¬φ∧ψ19→ψ20) ∧(¬φ∧ψ20→ψ21) ∧(¬φ∧ψ21→ψ22)
∧(¬φ∧ψ22→ψ23) ∧(¬φ∧ψ23→ψ24) ∧(¬φ∧ψ24→ψ25)
∧(¬φ∧ψ25→ψ26) ∧(¬φ∧ψ26→ψ27) ∧(¬φ∧ψ27→ψ28)
∧(¬φ∧ψ28→ψ29) ∧(¬φ∧ψ29→ψ30) ∧(¬φ∧ψ30→ψ31)
∧(¬φ∧ψ31→ψ32) ∧(¬φ∧ψ32→ψ33) ∧(¬φ∧ψ33→ψ34)
∧(¬φ∧ψ34→ψ35) ∧(¬φ∧ψ35→ψ36) ∧(¬φ∧ψ36→ψ37)
∧(¬φ∧ψ37→ψ38) ∧(¬φ∧ψ38→ψ39) ∧(¬φ∧ψ39→ψ40)
∧(¬φ∧ψ40→ψ41) ∧(¬φ∧ψ41→ψ42) ∧(¬φ∧ψ42→ψ43)
∧(¬φ∧ψ43→ψ44) ∧(¬φ∧ψ44→ψ45) ∧(¬φ∧ψ45→ψ46)
∧(¬φ∧ψ46→ψ47) ∧(¬φ∧ψ47→ψ48) ∧(¬φ∧ψ48→ψ49)
∧(¬φ∧ψ49→ψ50) ∧(¬φ∧¬ψ50→τ)

@korrelan: That was an amazing explanation, especially about multiplying the values in a lookup list to improve it. I think, that may help explain how it becomes increasingly challenging to squeeze more and more life likeness out of a chatbot design with computer programming.

@Don Patrick:  Please remind me of what you posted before. I was about to reply.

[Don Patrick]

I removed my post because it didn't add much constructive. To summarise:
A: associations. B: simple ideas typically underestimate the scope and detail of the problem.

[8pla.net]

Thank you Don. Yes, I remember applying your post to korrelan's post, about the extent to which lookup lists may be relevant to what A.I. deals with.  The scope presents one way to summarize, I think.

[DemonRaven]

Simple Chatbot Flowchart



Modeled and rendered using Blender. All original image from the Chatbot Tool Kit

Do you know who owns that website as http://www.chatbots.tk/  does not load for me so i am unsure if it still exists or not. It is listed under links.  They might want to check on it and correct  it.

[8pla.net]

Website http://www.chatbots.tk is an experiment for me to determina the pros and cons of running a chatbot on the internet at no cost.  One of the pros is that it is free. Yet, one of the cons is greater downtime for maintenance.   Free webhosts tend to host a very large number of websites which increases maintenance.

Perhaps you may have found one of the cons of free webhosting, DemonRaven?  One explanation may be that you attempted to load website http://www.chatbots.tk during a downtime for the free webhosting maintenance. 

For the purpose of linking to this experimental website, despite any increase in maintenance downtimes, it is intended to remain online indefinitely.  May I ask, concerning website http://www.chatbots.tk , has anyone else experienced downtime?

[spydaz]

children learn by environmental conditioning, (reinfrocment learning)....

Loads of research by (skinner, chomsky, bandura, paigot) etc...

Exactly how a AI is supposed to learn. !

people speak or choose words which are commonly used in thier environment(favorites)
People act the same as friends in thier environemt, copying accepted rules of behavior...
People know what the learn from thier environments... the world according to garf!

so if you teach an AI about the world it world will be the world it just learned from you.. just like a childs concept of right and wrong....

If you teach the ai to protect itself then if you attempt to destroy it then you may get destroyed. if you told it not to destroy humans then if a human attempted to destroy it , there would be a dilemma and after applying multiple scenarios, (nueral net) then it may come to the descision to break one of the two ideals ... let itself be destroyed (against programming) or destroy all humans(against programming) the results of the traning cases would prevail... as it must also be programmed to pick the best scenario ... (or try contriversal scenarios to see which sceanrio produced some positive reinforcement)

[DemonRaven]

Website http://www.chatbots.tk is an experiment for me to determina the pros and cons of running a chatbot on the internet at no cost.  One of the pros is that it is free. Yet, one of the cons is greater downtime for maintenance.   Free webhosts tend to host a very large number of websites which increases maintenance.

Perhaps you may have found one of the cons of free webhosting, DemonRaven?  One explanation may be that you attempted to load website http://www.chatbots.tk during a downtime for the free webhosting maintenance. 

For the purpose of linking to this experimental website, despite any increase in maintenance downtimes, it is intended to remain online indefinitely.  May I ask, concerning website http://www.chatbots.tk , has anyone else experienced downtime?

You were correct it is up now.