From Newsgroup: sci.logic

Hi,

Maybe I should write a blog post, titled
Introduction to AI Accelerator Prolog:

- specialized jobs -C-WAM (currently integerish stuff)
- -C-WAM uses no atomics, only comms
- -C-WAM uses warp, 30-40% more speed
- -C-WAM runs on GPU and CPU
- -C-WAM runs from within JavaScript, Python and Java

Feels like reinventing FGCS concurrent
logic programming.

LoL

Bye

Mild Shock schrieb:

Hi,

NVIDIA has just release RTX 3080 Mini.
Only the size of space bar, it easily
fits into a keyboard:

Nvidia RTX 3080 Mini! The Future of GPUs! https://www.instagram.com/p/C3gbuA8P0eE/

The association of logic programming has
coorperated with Morbid AI Inc. and used
a local GPT builder to bring Prolog

Expert Ginis on a keychain. You can now
easily carry around in your pocket:

Mini Hakan: Ask it anything about
constraint programming, contains the
wealth of CLP examples written in
different CLP dialect.

Mini Paul: Ask it anything about Jini
Prolog VMs. The complete hitchhiker guide
to engineering fabulous sequential
Prolog engines.

Mini Jan: Ask it anything about XPCE
and SWI. More than a manual , rather
a language monument. Fancy easter egg,
contains a complete GUI tracer.

Stay tuned, more to come...

Bye

Mild Shock schrieb:

Hi,

You just escaped AI dooms day. Humanity has
reset all internet and computers as a last resort
to prevent AGI developing, by an electromagnetic

pulse. You are stuck in G|+ttinger Wald and hunted
down a deer by your bare hands, the deer still
confused and tame because tourists were feeding it.

Now you have no knife, what do you do:

Chimpanzees Have Entered The Stone Age
https://www.youtube.com/watch?v=wPXX2I_uYjc

So we are just apes with internet.

Bye

Mild Shock schrieb:

Hi,

Ok I was looking at this learning challenge,
producing vector (y1,y2,y3,y4) from a vector
(x1,x2,x3,x4), System R can do it via least square?

| 0 0 0 1 |-a-a | x1 |-a-a-a-a | x4 |
| 0 0 1 0 |-a-a | x2 |-a =-a | x3 |
| 0 1 0 0 |-a-a | x3 |-a-a-a-a | x2 |
| 1 0 0 0 |-a-a | x4 |-a-a-a-a | x1 |

How it started:

"multiplicative RNNs arises naturally from a
proof-theoretic interpretation of next-token
prediction as nested intuitionistic implication"
Paul Tarau - 2026
https://arxiv.org/abs/2601.19915

How its going:

"Dave uses a PDP-11 to train a real Neural
Network complete with Transformers and
Attention so you can see them at their most basic."
Mr. Taskmanager - 2026
https://www.youtube.com/watch?v=OUE3FSIk46g

We see Doctor Frankstein in action from
the Bronze Age of Computing, producing
a Humunkulus, the progenitor of todays

Bulgakov Shuriks in the Hyperscale Age!

Bye

P.S.: My impression neither cut to the core, that
this incredible transformer most likely
produced this deterministic attention:

| -1 | * | k | + | 5 | = | k' |

Or differently expressed y_k = x_{5-k}.

How did the transformer do it? It produced
a neural network with 1216 parameters, but
didn't use embeddings or polar encoding

of positions. But if we strip the noise
and denoise from the position encoding,
the denoise is done via softmax. We somehow

must get the above, right? I still need to
verify my claim! BTW: The PDP-11 assembly
from 1979 uses wider example not with n=4

but with n=8.

--- Synchronet 3.22a-Linux NewsLink 1.2

From Newsgroup: sci.logic

Hi,

Somebody tasked the logic programming community
to make an +#-Prolog, i.e. have some alpha
conversion whatever builtin, plus maybe nominal
logic fresh variables whatever.

But who reads such nonsense in 2026:

Lifting E-Graphs: A Function Isn't a Constant
https://arxiv.org/abs/2606.22734

So you see where I am aiming with -C-WAM, surely
not +#-Prolog . I wouldn't care less about +#-Prolog.
-C is more fun than +#, and working on WAM is more
fun then creating some silly prototype of a

REPL of a homunkulus of a Prolog. -C beats +#:

- +# = abstract, academic, awkward
- -C = parallelism, performance, practical

Bye

Mild Shock schrieb:

Hi,

Maybe I should write a blog post, titled
Introduction to AI Accelerator Prolog:

- specialized jobs -C-WAM (currently integerish stuff)
- -C-WAM uses no atomics, only comms
- -C-WAM uses warp, 30-40% more speed
- -C-WAM runs on GPU and CPU
- -C-WAM runs from within JavaScript, Python and Java

Feels like reinventing FGCS concurrent
logic programming.

LoL

Bye

Mild Shock schrieb:

Hi,

NVIDIA has just release RTX 3080 Mini.
Only the size of space bar, it easily
fits into a keyboard:

Nvidia RTX 3080 Mini! The Future of GPUs!
https://www.instagram.com/p/C3gbuA8P0eE/

The association of logic programming has
coorperated with Morbid AI Inc. and used
a local GPT builder to bring Prolog

Expert Ginis on a keychain. You can now
easily carry around in your pocket:

Mini Hakan: Ask it anything about
constraint programming, contains the
wealth of CLP examples written in
different CLP dialect.

Mini Paul: Ask it anything about Jini
Prolog VMs. The complete hitchhiker guide
to engineering fabulous sequential
Prolog engines.

Mini Jan: Ask it anything about XPCE
and SWI. More than a manual , rather
a language monument. Fancy easter egg,
contains a complete GUI tracer.

Stay tuned, more to come...

Bye

Mild Shock schrieb:

Hi,

You just escaped AI dooms day. Humanity has
reset all internet and computers as a last resort
to prevent AGI developing, by an electromagnetic

pulse. You are stuck in G|+ttinger Wald and hunted
down a deer by your bare hands, the deer still
confused and tame because tourists were feeding it.

Now you have no knife, what do you do:

Chimpanzees Have Entered The Stone Age
https://www.youtube.com/watch?v=wPXX2I_uYjc

So we are just apes with internet.

Bye

Mild Shock schrieb:

Hi,

Ok I was looking at this learning challenge,
producing vector (y1,y2,y3,y4) from a vector
(x1,x2,x3,x4), System R can do it via least square?

| 0 0 0 1 |-a-a | x1 |-a-a-a-a | x4 |
| 0 0 1 0 |-a-a | x2 |-a =-a | x3 |
| 0 1 0 0 |-a-a | x3 |-a-a-a-a | x2 |
| 1 0 0 0 |-a-a | x4 |-a-a-a-a | x1 |

How it started:

"multiplicative RNNs arises naturally from a
proof-theoretic interpretation of next-token
prediction as nested intuitionistic implication"
Paul Tarau - 2026
https://arxiv.org/abs/2601.19915

How its going:

"Dave uses a PDP-11 to train a real Neural
Network complete with Transformers and
Attention so you can see them at their most basic."
Mr. Taskmanager - 2026
https://www.youtube.com/watch?v=OUE3FSIk46g

We see Doctor Frankstein in action from
the Bronze Age of Computing, producing
a Humunkulus, the progenitor of todays

Bulgakov Shuriks in the Hyperscale Age!

Bye

P.S.: My impression neither cut to the core, that
this incredible transformer most likely
produced this deterministic attention:

| -1 | * | k | + | 5 | = | k' |

Or differently expressed y_k = x_{5-k}.

How did the transformer do it? It produced
a neural network with 1216 parameters, but
didn't use embeddings or polar encoding

of positions. But if we strip the noise
and denoise from the position encoding,
the denoise is done via softmax. We somehow

must get the above, right? I still need to
verify my claim! BTW: The PDP-11 assembly
from 1979 uses wider example not with n=4

but with n=8.

--- Synchronet 3.22a-Linux NewsLink 1.2

From Newsgroup: sci.logic

On 7/17/2026 2:16 AM, Mild Shock wrote:

Hi,

Maybe I should write a blog post, titled
Introduction to AI Accelerator Prolog:

- specialized jobs -C-WAM (currently integerish stuff)
- -C-WAM uses no atomics, only comms
- -C-WAM uses warp, 30-40% more speed
- -C-WAM runs on GPU and CPU
- -C-WAM runs from within JavaScript, Python and Java

[...]

No atomic fetch-and-add?
--- Synchronet 3.22a-Linux NewsLink 1.2

From Newsgroup: sci.logic

Hi,

The pi in pi-WAM refers to pi-calculus.
pi-calculus has not atomic(i32).

The -C-calculus is a universal model of computation.
This was first observed by Milner in his paper
"Functions as Processes",[10] in which he presents
two encodings of the lambda-calculus in the -C-calculus. https://en.wikipedia.org/wiki/%CE%A0-calculus

LoL

Bye

Chris M. Thomasson schrieb:

On 7/17/2026 2:16 AM, Mild Shock wrote:

Hi,

Maybe I should write a blog post, titled
Introduction to AI Accelerator Prolog:

- specialized jobs -C-WAM (currently integerish stuff)
- -C-WAM uses no atomics, only comms
- -C-WAM uses warp, 30-40% more speed
- -C-WAM runs on GPU and CPU
- -C-WAM runs from within JavaScript, Python and Java

[...]

No atomic fetch-and-add?

--- Synchronet 3.22a-Linux NewsLink 1.2

From Newsgroup: sci.logic

Hi,

Milners fickle() is here:

Functions as processes
https://inria.hal.science/inria-00075405

After Theorem 7.7:

So in P we construct a fickle rCyfunctionrCO which
behaves differently on successive calls.

Here is a pi-WAM run in Dogelog Player, using the emulator:

Dogelog Spieler 2.2.4, Oracle Corporation, Java 26.0.1
(c) 1985-2026, XLOG Technologies AG, Schweiz
?- ensure_loaded(library(edge/brainfog)).
true.
?- emulate((between(1,2,Y),in(X),out(Y))).
: 0
1
: 0
2
fail.

The emulator is portable, can be run every Prolog
system. But it is only 1 process. So its better
to use the n process backends for CPU or GPU.

Which are less portable, not anymore pure Prolog,
a great deal of thread start and join infrastructure
as well, and a native Hack VM.

The comms across process is not yet implemented.
But the in/1 and out/1 instructions are already
there. But they currently go to stdin/stdout.

Bye

Mild Shock schrieb:

Hi,

The pi in pi-WAM refers to pi-calculus.
pi-calculus has not atomic(i32).

The -C-calculus is a universal model of computation.
This was first observed by Milner in his paper
"Functions as Processes",[10] in which he presents
two encodings of the lambda-calculus in the -C-calculus. https://en.wikipedia.org/wiki/%CE%A0-calculus

LoL

Bye

Chris M. Thomasson schrieb:

On 7/17/2026 2:16 AM, Mild Shock wrote:

Hi,

Maybe I should write a blog post, titled
Introduction to AI Accelerator Prolog:

- specialized jobs -C-WAM (currently integerish stuff)
- -C-WAM uses no atomics, only comms
- -C-WAM uses warp, 30-40% more speed
- -C-WAM runs on GPU and CPU
- -C-WAM runs from within JavaScript, Python and Java

[...]

No atomic fetch-and-add?

--- Synchronet 3.22a-Linux NewsLink 1.2

From Newsgroup: sci.logic

On 07/17/2026 04:50 PM, Mild Shock wrote:

Hi,

Milners fickle() is here:

Functions as processes
https://inria.hal.science/inria-00075405

After Theorem 7.7:

So in P we construct a fickle rCyfunctionrCO which
behaves differently on successive calls.

Here is a pi-WAM run in Dogelog Player, using the emulator:

Dogelog Spieler 2.2.4, Oracle Corporation, Java 26.0.1
(c) 1985-2026, XLOG Technologies AG, Schweiz
?- ensure_loaded(library(edge/brainfog)).
true.
?- emulate((between(1,2,Y),in(X),out(Y))).
: 0
1
: 0
2
fail.

The emulator is portable, can be run every Prolog
system. But it is only 1 process. So its better
to use the n process backends for CPU or GPU.

Which are less portable, not anymore pure Prolog,
a great deal of thread start and join infrastructure
as well, and a native Hack VM.

The comms across process is not yet implemented.
But the in/1 and out/1 instructions are already
there. But they currently go to stdin/stdout.

Bye

Mild Shock schrieb:

Hi,

The pi in pi-WAM refers to pi-calculus.
pi-calculus has not atomic(i32).

The -C-calculus is a universal model of computation.
This was first observed by Milner in his paper
"Functions as Processes",[10] in which he presents
two encodings of the lambda-calculus in the -C-calculus.
https://en.wikipedia.org/wiki/%CE%A0-calculus

LoL

Bye

Chris M. Thomasson schrieb:

On 7/17/2026 2:16 AM, Mild Shock wrote:

Hi,

Maybe I should write a blog post, titled
Introduction to AI Accelerator Prolog:

- specialized jobs -C-WAM (currently integerish stuff)
- -C-WAM uses no atomics, only comms
- -C-WAM uses warp, 30-40% more speed
- -C-WAM runs on GPU and CPU
- -C-WAM runs from within JavaScript, Python and Java

[...]

No atomic fetch-and-add?

So he partial-computes his alpha to alpha_alpha
in his Futamura projection, yet, that's partial computation,
why not make Minsky counter-machines using the bottom-up
approach of making arithmetizations and DFA's.


Of course the lambda-calculus and pi-calculus are great
things for models of types and communicating sequential processes.

Roberto di Cosmo has a great book on types, and more than one.

https://books.google.com/books/about/Isomorphisms_of_Types.html?id=cdJZRjIxavwC


--- Synchronet 3.22a-Linux NewsLink 1.2

From Newsgroup: sci.logic

Hi,

Mild Shock schrieb:

The -C-calculus is a universal model of computation.
This was first observed by Milner in his paper
"Functions as Processes",[10] in which he presents
two encodings of the lambda-calculus in the -C-calculus. https://en.wikipedia.org/wiki/%CE%A0-calculus

Ross Finlayson schrieb:

Of course the lambda-calculus and pi-calculus
are great things for models of types and communicating
sequential processes.

Roberto di Cosmo has a great book on types,
and more than one.

https://books.google.com/books/about/Isomorphisms_of_Types.html?id=cdJZRjIxavwC


When a 1970s paper claims a relation ship between
pi-calculus and lambda calculus, then both calculi
refer to a typeless calculi.

Types are a later invention. The original lambda
calculus was typeless. Church encodings came later,
but for example the Church Turing hypotheses is

formulated along typeless lambda calculus.

Bye

P.S.: Prolog is also typeless. I do not intend to
add any types to pi-WAM either.

Ross Finlayson schrieb:

On 07/17/2026 04:50 PM, Mild Shock wrote:

Hi,

Milners fickle() is here:

Functions as processes
https://inria.hal.science/inria-00075405

After Theorem 7.7:

So in P we construct a fickle rCyfunctionrCO which
behaves differently on successive calls.

Here is a pi-WAM run in Dogelog Player, using the emulator:

Dogelog Spieler 2.2.4, Oracle Corporation, Java 26.0.1
(c) 1985-2026, XLOG Technologies AG, Schweiz
?- ensure_loaded(library(edge/brainfog)).
true.
?- emulate((between(1,2,Y),in(X),out(Y))).
: 0
1
: 0
2
fail.

The emulator is portable, can be run every Prolog
system. But it is only 1 process. So its better
to use the n process backends for CPU or GPU.

Which are less portable, not anymore pure Prolog,
a great deal of thread start and join infrastructure
as well, and a native Hack VM.

The comms across process is not yet implemented.
But the in/1 and out/1 instructions are already
there. But they currently go to stdin/stdout.

Bye

Mild Shock schrieb:

Hi,

The pi in pi-WAM refers to pi-calculus.
pi-calculus has not atomic(i32).

The -C-calculus is a universal model of computation.
This was first observed by Milner in his paper
"Functions as Processes",[10] in which he presents
two encodings of the lambda-calculus in the -C-calculus.
https://en.wikipedia.org/wiki/%CE%A0-calculus

LoL

Bye

Chris M. Thomasson schrieb:

On 7/17/2026 2:16 AM, Mild Shock wrote:

Hi,

Maybe I should write a blog post, titled
Introduction to AI Accelerator Prolog:

- specialized jobs -C-WAM (currently integerish stuff)
- -C-WAM uses no atomics, only comms
- -C-WAM uses warp, 30-40% more speed
- -C-WAM runs on GPU and CPU
- -C-WAM runs from within JavaScript, Python and Java

[...]

No atomic fetch-and-add?

So he partial-computes his alpha to alpha_alpha
in his Futamura projection, yet, that's partial computation,
why not make Minsky counter-machines using the bottom-up
approach of making arithmetizations and DFA's.

Of course the lambda-calculus and pi-calculus are great
things for models of types and communicating sequential processes.

Roberto di Cosmo has a great book on types, and more than one.

https://books.google.com/books/about/Isomorphisms_of_Types.html?id=cdJZRjIxavwC

--- Synchronet 3.22a-Linux NewsLink 1.2

From Newsgroup: sci.logic

Hi,

Ha Ha, TLo @ FOM schrieb:

Lastly--there is the Church-Turing-thesis itself.
I cannot imagine what form a proof could have. But at
least a refutation is conceivable, if extremely far-fetched.

Can the Church Turing hypotheses be refuted? Well
since it is stricly speaking only a hypothesis and
not a thesis, it cannot be that much refuted,

since it is a kind of exhaustion argument defining
a category by examples. The history goes as follows:

In 1935rCo36,[7] Alonzo Church formalized the
concept of effectively calculable functions by proposing
that they are general recursive functions, or,
equivalently, ++-definable functions.

https://en.wikipedia.org/wiki/Church%E2%80%93Turing_thesis

Does his ++-definable include types? Nope. Do
we need large cardinal axioms? It depends.
Church ++-computable was syntactically defined,

it didn't have a semantic referent. The search
for a semantic referent gave rise to certain
additional problems of set theory and type theory.

Have Fun!

Bye

Mild Shock schrieb:

Hi,

Mild Shock schrieb:

The -C-calculus is a universal model of computation.
This was first observed by Milner in his paper
"Functions as Processes",[10] in which he presents
two encodings of the lambda-calculus in the -C-calculus. https://en.wikipedia.org/wiki/%CE%A0-calculus

Ross Finlayson schrieb:

Of course the lambda-calculus and pi-calculus
are great things for models of types and communicating
sequential processes.

Roberto di Cosmo has a great book on types,
and more than one.

https://books.google.com/books/about/Isomorphisms_of_Types.html?id=cdJZRjIxavwC

When a 1970s paper claims a relation ship between
pi-calculus and lambda calculus, then both calculi
refer to a typeless calculi.

Types are a later invention. The original lambda
calculus was typeless. Church encodings came later,
but for example the Church Turing hypotheses is

formulated along typeless lambda calculus.

Bye

P.S.: Prolog is also typeless. I do not intend to
add any types to pi-WAM either.

Ross Finlayson schrieb:

On 07/17/2026 04:50 PM, Mild Shock wrote:

Hi,

Milners fickle() is here:

Functions as processes
https://inria.hal.science/inria-00075405

After Theorem 7.7:

So in P we construct a fickle rCyfunctionrCO which
behaves differently on successive calls.

Here is a pi-WAM run in Dogelog Player, using the emulator:

Dogelog Spieler 2.2.4, Oracle Corporation, Java 26.0.1
(c) 1985-2026, XLOG Technologies AG, Schweiz
?- ensure_loaded(library(edge/brainfog)).
true.
?- emulate((between(1,2,Y),in(X),out(Y))).
: 0
1
: 0
2
fail.

The emulator is portable, can be run every Prolog
system. But it is only 1 process. So its better
to use the n process backends for CPU or GPU.

Which are less portable, not anymore pure Prolog,
a great deal of thread start and join infrastructure
as well, and a native Hack VM.

The comms across process is not yet implemented.
But the in/1 and out/1 instructions are already
there. But they currently go to stdin/stdout.

Bye

Mild Shock schrieb:

Hi,

The pi in pi-WAM refers to pi-calculus.
pi-calculus has not atomic(i32).

The -C-calculus is a universal model of computation.
This was first observed by Milner in his paper
"Functions as Processes",[10] in which he presents
two encodings of the lambda-calculus in the -C-calculus.
https://en.wikipedia.org/wiki/%CE%A0-calculus

LoL

Bye

Chris M. Thomasson schrieb:

On 7/17/2026 2:16 AM, Mild Shock wrote:

Hi,

Maybe I should write a blog post, titled
Introduction to AI Accelerator Prolog:

- specialized jobs -C-WAM (currently integerish stuff)
- -C-WAM uses no atomics, only comms
- -C-WAM uses warp, 30-40% more speed
- -C-WAM runs on GPU and CPU
- -C-WAM runs from within JavaScript, Python and Java

[...]

No atomic fetch-and-add?

So he partial-computes his alpha to alpha_alpha
in his Futamura projection, yet, that's partial computation,
why not make Minsky counter-machines using the bottom-up
approach of making arithmetizations and DFA's.


Of course the lambda-calculus and pi-calculus are great
things for models of types and communicating sequential processes.

Roberto di Cosmo has a great book on types, and more than one.

https://books.google.com/books/about/Isomorphisms_of_Types.html?id=cdJZRjIxavwC

--- Synchronet 3.22a-Linux NewsLink 1.2