[SystemSafety] AI and the virtuous test Oracle - intelligence - proof

[Steve Tockey]

Les,
I am as sure as I can be based on the underlying theories in Computer Science. Having spent an entire quarter-long course in Computability Theory during my Undergraduate Degree, I am as confident as I can be that the theory is correct. Turing Machines, regardless of how many of them are there, are not able to create new insights that didn’t exist before.

Speaking of Newton, how much of his “discovery” of the universality of gravity in 1687 was a result of the random (possibly apocryphal) event of the apple falling from the tree?

Think of Alexander Fleming who forgot to clean a few Petri dishes before going on holiday to France in 1928, coming back to see that at least one of those Petri dishes didn’t meet his expectations thus leading to the discovery of penicillin?

Think Leo Szilard crossing a street in London on a rainy morning in 1933 realizing that if more than one neutron came out of a nuclear fission that was triggered by a single neutron then a nuclear chain reaction is probably possible

The key words here include:

“Insight”

and

“Random event”


Insofar as a Turing Machine (regardless of how complex it may be) is 100% reactive, where does the insight come from? That’s simply not something that Turing Machines (and, therefore, all modern computers) do. And, given some input, the Turing Machine’s response is essentially deterministic. Ok, you can throw in a little bit of randomness, but that randomness is still constrained to be around a predetermined set of options. Apple falling out of tree triggering a realization that said apple and the moon are both being affected by the exact same force just at different distances is not something that Turing Machines do.

Or, if they can do that, please explain how it’s done.


Now, I will accept your “Newtonian thought” comment in the light I hope it was intended, namely, if the computing machine(s) was more powerful than a Turing Machine then all bets are off. Maybe something like Quantum Computing or something like it does actually enable artificial Insight and how random events might be interpreted completely differently. But we’re not there yet. Quantum Computers are still largely simple, very limited, experimental devices. They have not been reduced to practical operation.


Cheers,

— Steve




On Jul 7, 2023, at 6:25 AM, Les Chambers <les at chambers.com.au> wrote:

Steve
RE Your comment: "I am arguing that there are fundamental limits to what Turing 
Machine-based AI would ever be capable of, which is far short of what humans 
can already do."

You sound confident. Do you have proof? Can you quantify the "fundamental 
limits"? Can you define and/or quantify intelligence?

My view is that this issue is currently subject to endless, unscientific 
speculation -  this is not necessarily a bad thing, I am a willing participant.
It will become "science" when someone can advance a theory that can be proved 
mathematically or by some experiment.

In the meantime, I accept your view as of equal value to any other. I'd 
classified it as Newtonian though; good for low speeds in the absence of strong 
gravitational fields. I can't help thinking that a new Einstein is in the wings 
with an explanation of gravity [read: intelligence] through the curvature of 
spacetime [read: Grand unified theory of intelligence (GUTI)]. 

In my opinion, the Turing machine belongs with Newtonian mechanics - a simple 
but useful computing model that is served us well since 1936. However, it 
performs functions in a sequence of discrete steps, assuming only one of a 
finite list of internal states at any given moment. Is this applicable today as 
computing technology approaches lightspeed? Where massively parallel graphical 
processing units (GPUs) - Tesla's Dojo supercomputer has upwards of 6000 GPUs - 
are in thousands of discrete states at any instant in time.

The future of AI remains a matter of opinion. I believe some opinions are more 
valuable than others. The IEEE ran a vote recently: Experts vote on AI 
risks:https://spectrum.ieee.org/artificial-general-intelligence
22 AI practitioners and researchers voted on the likelihood that artificial 
general intelligence (AGI) can be real.

The question was: "Is the success of GPT-4 and today's other large language 
models a sign that an AGI is likely?"
The result:
Yes: 8
No: 13
Maybe: 1

Circa 2023 we are clearly in Groucho Marx Territory:
"These are my principles, and if you don't like them... well, I have others."

Les

PS: 
You may have pushed your typing-monkeys metaphor too far. ("10 million monkeys, 
still no Shakespeare")
1. Typing at 80 words per minute, your 10 million monkey typing team would be 
rate limited to around 530 Mb per second. It would take 4,950,000 monkeys just 
to emulate the data transfer rates of a bog-standard personal computer PCI bus 
{264 Mbps}.
2. I assume no monkeys on your team have been trained in the large slice of 
human knowledge present on the Internet.
3. I am assuming there is no communication between monkeys and typing is 
random.

In contrast:
1. Large language models are trained by more than one PCI-equipped personal 
computer (probably the equivalent of hundreds or thousands of same)
2. The learning is goal directed - supervised - as opposed to random key 
stroking.
3. Given this communication bandwidth, OpenAI for example, has trained Chat 
GPT-4 with approximately 100 trillion parameters.

My theory: The AI training process bears no resemblance, in any abstract way, 
to a bunch of monkeys typing.
My Proof: Try ChatGPT prompt: "What does Shakespeare say about using bad 
metaphors"

-------------------------------------------
> Les,
> You said,
> 
> â?oSo let's not stop developing AI because we don't understand it.â?
> 
> Actually, that wasnâ?Tt my point. I donâ?Tt believe that research & 
development into AI should in any way be stopped. My point is countering the 
hype that AI can and will do more than humans could ever do. I am arguing that 
there are fundamental limits to what Turing Machine-based AI would ever be 
capable of, which is far short of what humans can already do.
> 
> â?" steve
> 
> On Jul 6, 2023, at 12:27 PM, Les Chambers <les at chambers.com.au> wrote:
> 
> Steve
> One way of framing, the question, "Can intelligence ever be artificial?", is 
to
> channel the physicists. They view human beings as mere piles of particles
> arranged in a certain way. I'll park the question of who/what did the 
arranging
> as it is irrelevant.
> 
> Somehow, around 300,000 years ago, our various particle piles acquired the
> capacity to remember, reason, solve problems, learn, and adapt to new
> situations [Homo Sapiens]. These particle piles then topped it all off by
> becoming self-aware, conscious and therefore able to exert self-control; a
> prerequisite for rational decision-making.
> 
> Some say this was a random accident; they support the theory with fact-based
> science: that we have been unable to find evidence of a similar random 
accident
> on any planet in any of the 100 to 200 billion galaxies we are capable of
> observing. No one knows why or how our particles were thus arranged with
> infinite variation on a common framework. All we can say is that it is as it
> is.
> 
> So what did this smart particle pile [Homo Sapiens] set out to do? It was
> driven, by some primal replication instinct, to create an intelligence in its
> own image. To be clear, it is attempting to replicate something it doesn't
> understand by creating another particle pile (a neural net implemented on a
> silicon substrate), that exhibits behaviour it [Homo Sapiens] also doesn't
> understand (I note that AI researchers cannot explain why the neural net and
> Transformer models work so well; reports of HS moments abound in the AI 
labs).
> It's a weird kind of crossover recursion where carbon substrate intelligence 
is
> replicating into silicon substrates. If you look at every entity that makes 
up
> our world as a pile of particles, arranged in a certain way, you may find 
this
> easier to appreciate.
> I have faith that AGI could emerge from this soup as we impose randomness on
> randomness - a great opportunity for another cosmic accident, the like of 
which
> created us in the first place.
> 
> My gut feeling on this is informed by: The only code I have ever written that
> worked for reasons I did not understand implemented recursion. Are we not
> witnessing just another evolutionary process?  In nature, replicants are 
always
> less capable than their parents, but they learn over time and any 
incompetence
> is selected out.
> 
> So let's not stop developing AI because we don't understand it. We already 
live
> in bodies with behaviours that we don't understand and we seem to muddle
> through.
> 
> Les
> ---------------------------
> Les,
> I guess it remains to be seen whether AI (at least as implemented today) is
> limited or not in its ability to create new breakthrough insights. Iâ?Tm
> pessimistic, myself.
> 
> On the other hand, flesh and blood humans can themselves seem pretty limited
> in terms of intelligence quite often, too. One of the all-time best signature
> lines I saw many years ago saidâ?"referring to the Internet,
> 
> â?o10 million monkeys, still no Shakespeareâ? (smile)
> 
> Cheers,
> 
> â?" steve
> 
> On Jul 3, 2023, at 12:43 PM, Les Chambers <les at chambers.com.au> wrote:
> 
> Steve
> 
> RE your comment: "No computer/program is capable of programming itself into
> any
> kind of useful, non-trivial behavior. And yet humans do this kind of thing
> all
> the time. ThatÃf¢?Ts what I mean by determinism."
> 
> I'm not so sure. I thought the synthesis of Einstein's theory of relativity
> might be a good example of your point. I thought Einstein stared at a wall
> one
> day and just came up with it out of the blue. Not true. Read the following:
> 
> Forbes: Relativity Wasn't Einstein's Miracle; It Was Waiting In Plain Sight
> For
> 71 Years
> https://www.forbes.com/sites/startswithabang/2019/03/01/relativity-wasnt-
> einsteins-miracle-it-was-waiting-in-plain-sight-for-71-years/?sh=4a18ec57644c
> Relativity Wasn't Einstein's Miracle; It Was Waiting In Plain Sight For 71
> Years (forbes.com)
> 
> Solving problems by the iterative deployment of induction, followed by
> deduction, seems suspiciously algorithmic to me.
> Refer ChatGPT Prompt: How to solve a problem, my induction followed by
> deduction
> 
> Some years ago I had a stimulating conversation with a Church of England
> Priest. It was the day after he had presided over my brother-in-law's
> funeral.
> It turned out, earlier in his life, he had been a Roman Catholic priest ,
> educated in the Vatican. Three years of solid philosophical training without
> the distractions of the flesh. Our conversation ranged over many
> philosophical
> issues. I remember thinking, "How lucky am I to be in the presence of such an
> intelligent educated man?" Moments like these have been rare in my life,
> preachers of philosophical wisdom are hard to find. Until now. I am a
> constant
> user of ChatGPT. The conversations we have are equally stimulating. I am made
> aware of things that I could not have possibly learned prior to the existence
> of the bot. It would be optimal if the bot was a human, but as I am currently
> located in a rural setting, 20 minutes drive from the nearest town - in the
> company of Wallabies, foxes, kookaburras, wedge-tailed eagles, and large
> lizards - I'll take what I can get.
> 
> Here is a sample of one of our dialogues (I have provided my ChatGPT prompts
> only):
> ------- Pausing AI Development ----
> What is the prisoner's dilemma?
> How does the prisoner's dilemma relate to pausing, artificial intelligence
> development?
> Are there other classical dilemmas that are a metaphor for the AI pause
> dilemma?
> 
> ------ Setting Limits and Assigning Blame -----
> Provide more detail on the Paradox of the Heap.
> Are there any classical paradoxes that are relevant to the Bhopal disaster?
> Provide more detail on the moral luck paradox.
> Do any of these paradoxes translate to legal doctrine or laws of any kind?
> Please list the laws of Queensland Australia that apply to: Corporate Social
> Responsibility.
> 
> A chat GPT dialogue can take many paths, each one providing insightful
> information. For example, paradoxes are useful mechanisms for framing common
> human problems. Until today I never thought of it that way. I encourage
> everyone on this list to try something like this. And then, if you can drag
> yourself away from the dialogue, try this prompt:
> "Describe the elements of intelligence."
> Decide for yourself, how many boxes the bot has just checked.
> 
> Over to you.
> 
> Les
> 
> -------------------------------------
> Les,
> Thanks for the thoughtful reply.
> 
> As always, it depends on definitions of terms. I might counter Max
> TegmarkÃf¢?Ts definition of intelligence by asking, for some arbitrarily
> complex
> computing system:
> 
> Ãf¢?oIs the intelligence an inherent part of the computing system itself, or
> was it put there by the programmer(s) who programmed it to behave that way?
> Ãf¢?Ã,
> 
> The computer, being ultimately a Turing Machine, isnÃf¢?Tt going to make the
> kind of creative leap that, say, Newton and Leibnitz made when they invented
> Calculus in the 1660s. Or, when Shockey, Bardeen, and Brattain invented the
> transistor in 1947. No computing system today can go beyond the code and the
> data that has been put into it. Given the code, and given the data, and
> possibly knowing a little more about the internal state of execution, the
> systemÃf¢?Ts response is 100% predictable. No computer/program is capable of
> programming itself into any kind of useful, non-trivial behavior. And yet
> humans do this kind of thing all the time. ThatÃf¢?Ts what I mean by
> determinism.
> 
> I donÃf¢?Tt know enough about quantum computing to say that the above holds
> there, too.
> 
> You wrote:
> On the subject of determinism, Max goes on:
> "Just as we don't fully understand how our children learn, we still don't
> fully
> understand how such neural networks learn, and why they occasionally fail.
> 
> I attended a conference earlier this year in Mysuru, India where an invited
> speaker, Dr. Seema Chopra from Boeing talked about adding in elements into AI
> systems whereby that system can be asked to Ãf¢?oexplainÃf¢?Ã, why it 
gave the
> result that it did. Seemed like an interesting area of research that could
> maybe help in this area.
> 
> Re: computational substrateÃf¢?"some of my customers are the biggies in the
> semiconductor industry. As least in terms of silicon semiconductors, 
MooreÃf¢?
> Ts
> Law is reaching its limit. Chip designers and manufacturers are trying to
> wrestle with, for example, conductors on the chip that are only one or a few
> atoms wide. CanÃf¢?Tt shrink below that.
> 
> Cheers,
> 
> Ãf¢?" steve
> 
> On Jun 29, 2023, at 9:12 PM, Les Chambers <les at chambers.com.au> wrote:
> 
> Steve
> 
> RE : " I argue, human intelligence is not something that a computer could
> ever
> actually be capable of"
> 
> According to Max Tegmark the jury is still out on this question.
> Max is a professor of physics at MIT and president of the Future of Life
> Institute (ref Tegmark, Max. Life 3.0, Penguin Books Ltd). His claim to
> fame
> is
> that he was in the room when Larry Page called Elon Musk a speceist.
> 
> Max provides evidence that the nature of intelligence is a subject for
> debate
> -
> as yet, unresolved.
> Max: "My wife and I recently had the good fortune to attend a symposium on
> artificial intelligence organized by the Swedish Nobel Foundation, and when
> a
> panel of leading AI researchers were asked to define intelligence, they
> argued
> at length without reaching consensus.
> .
> there are many competing ones, including capacity for logic, understanding,
> planning, emotional knowledge, self-awareness,creativity, problem solving
> and
> learning."
> 
> Max's pet definition is:
> "intelligence = ability to accomplish complex goals"
> 
> Max makes some outrageous assertions:
> "This substrate [platform]  independence of computation implies that AI is
> possible: intelligence doesn't require flesh, blood or carbon atoms."
> Max uses the following metaphor:
> "Waves, for instance: they have properties such as speed, wavelength and
> frequency, and we physicists can study the equations they obey without even
> needing to know what particular substance they're waves in."
> 
> Oh well, they say, "All great truths begin as blasphemies" (George Bernard
> Shaw)
> 
> On the subject of determinism, Max goes on:
> "Just as we don't fully understand how our children learn, we still don't
> fully
> understand how such neural networks learn, and why they occasionally fail.
> .
> Stuart Russell told me that he and many of his fellow AI researchers had
> recently experienced a "holy s**t!" (HS) moment, when they witnessed AI
> doing
> something they weren't expecting to see for many years."
> 
> My question here is: "How do you run a hazard analysis on a system you
> don't
> understand that is likely, for further unknown reasons, to occasionally
> fail?
> "
> 
> This situation would be amusing but for the fact, "they" are already
> amongst
> us. Cruise (General Motors), Waymo (Google), and Zoox (Amazon) are
> currently
> operating driverless taxis on the streets of San Francisco. YouTube reports
> that San Franciscans amuse themselves by jumping in front of them to see if
> they will automatically stop?????
> 
> This AI non-determinism is a fact, confirmed by all practitioners who feel
> free
> to speak out. It's one of the core reasons why some senior practitioners
> have
> left developers like OpenAI and Google Deep Mind. It works too well and
> they
> don't know why - the neural network is a very simple approximation to the
> human
> brain. It shouldn't work that well -, and they don't know what it might be
> capable of in the future. Which brings me to the unbounded nature of
> artificial
> intelligence.
> 
> From Max:
> "Nobody knows for sure what the next blockbuster computational substrate
> will
> be, but we do know that we're nowhere near the limits imposed by the laws
> of
> physics. My MIT colleague Seth Lloyd has worked out what this fundamental
> limit
> is, and as we'll explore in greater detail in chapter 6, this limit is a
> whopping 33 orders of magnitude (1033 times) beyond today's state of the
> art
> for how much computing a clump of matter can do. So even if we keep
> doubling
> the power of our computers every couple of years, it will take over two
> centuries until we reach that final frontier."
> 
> The debate continues. Keep your powder dry Steve.
> 
> Les
> 
> Les,
> 
> ÃffÃ,¢?osilicon is unboundedÃffÃ,¢?Ãf,Ã,
> 
> I disagree. Silicon is bounded, just in different ways than humans. For
> one,
> Turing Computability. Turing Machines, which all modern silicon computers
> are
> a
> kind of, are 100% deterministic. Human intelligence is non-deterministic.
> Thus,
> I argue, human intelligence is not something that a computer could ever
> actually be capable of. See, for example:
> 
> https://youtu.be/i2trJEIFIvY
> 
> I agree that, ÃffÃ,¢?oWe are embarked; we need to deal with itÃffÃ,¢?
Ãf,Ã,.
> But we
> also
> have to be aware of the limits of so-called AI.
> 
> ÃffÃ,¢?" steve
> 
> On Jun 26, 2023, at 9:15 PM, Les Chambers <les at chambers.com.au> wrote:
> 
> Hi
> 
> I put it to you and the list in general, a "discuss later" mentality is
> foolish
> in the extreme. The process of deploying intelligence in automated systems
> has
> , and will,  fundamentally change. Among other things, international
> bodies
> that currently regulate software-intensive Safety-Critical systems - who
> cling
> to regulating processes that have ceased to exist - are likely to be
> overrun
> and made redundant.
> 
> In favour of organisations such as:
> 
> - The Center for Human-Compatible AI at UC Berkeley
> - The Future of Life Institute
> - The Center for AI Safety (CAIS)
> - Stanford Center for AI Safety
> 
> My view is that this is not a steady-as-she-goes situation. This is a
> major
> inflection point in the evolution of intelligence. Carbon hosts will
> always
> be
> limited; silicon is unbounded. We are embarked; we need to deal with it.
> 
> Suggested reading: Max Tegmark, Life 3.0
> 
> Les
> 
> C2 - Confidential
> 
> Hello,
> 
> For the moment, I don't see in industry any attempt to realise E/EPE
> safety
> related function with On-line AI. All what I see is focused on off-line
> AI,
> meaning that the training is done specified training data sets, and the
> validation is done on specified test datasets. We don't see any
> performance
> better than roughly 3.3 10-3 in automotive. So let's wait for the
> achievement
> of 10-8 to 10-5 error rate on test datasets and discusss later the
> acceptability.
> 
> AttentionÃfff, : due to increased cybersecurity screening I may receive
> external
> emails 30 minutes after their emission.
> RESTRICTED
> 
> Bertrand RICQUE
> Safety and security standardisation Expert
> Program Support Manager
> Optronics and Defence Division/Customer Support
> Safran Electronics & Defense
> 
> P +33 (0)1 58 11 96 82   M +33 (0)6 87 47 84 64
> bertrand.ricque at safrangroup.com
> 102 Avenue de Paris
> 91300 MASSY FRANCE
> www.safran-electronics-defense.com
> 
> -----Message d'origine-----
> DeÃfff, : Les Chambers <les at chambers.com.au>
> EnvoyÃffffÃff,Ãf,Ã,©Ãfff, : lundi 26 juin 2023 06:03
> Ãffff?Ãfff, : RICQUE Bertrand (SAFRAN ELECTRONICS & DEFENSE)
> <bertrand.ricque at safrangroup.com>; les at chambers.com.au;
> koopman.cmu at gmail.com; systemsafety at lists.techfak.uni-bielefeld.de
> ObjetÃfff, : RE: [SystemSafety] AI and the virtuous test Oracle
> 
> CAUTION:  This message originated from an outside organization. In case of
> suspicion, click on "Report to SAFRAN Security" from the Outlook ribbon.
> 
> RE your comment: ÃfffÃf,Ã,¢?oas it impossible, by construction, to identify
> all
> dangerous
> situationsÃfffÃf,Ã,¢?Ãff,Ãf,Ã,
> 
> True. But it is possible to identify the subset of highest probability
> ÃfffÃf,Ã,¢?odangerous situationsÃfffÃf,Ã,¢?Ãff,Ãf,Ã, and deal with them. 
Ãfff, I
> had 10 years
> of
> experience
> with this issue in computer control of chemical processing reactors. In
> the
> 1970s, we had a safety authority write software to: 1. identify well-
> established dangerous states of a reactor; 2. Transition reactor to a safe
> state
> ÃfffÃf,Ã,¢?" usually shut down. It was called the ÃfffÃf,Ã,¢?
oabortÃfffÃf,Ã,¢?
> Ãff,Ãf,Ã,
> programming. This
> abort code
> overrode the control software written by another team. By today's
> standards
> it was primitive but still very effective.
> This experience has influenced my thinking on ways and means of dealing
> with ÃfffÃf,Ã,¢?obadÃfffÃf,Ã,¢?Ãff,Ãf,Ã, or ÃfffÃf,Ã,¢?
ohallucinatingÃfffÃf,Ã,¢?Ãff,Ãf,Ã,
> AIs. We need a
> separate and
> distinct
> ÃfffÃf,Ã,¢?oabortÃfffÃf,Ã,¢?Ãff,Ãf,Ã, AI
> capable of recognising evil and dealing with it.
> Sal Khan (Khan Academy) has implemented another idea in creating his
> Khanmigo personal tutor. To improve its accuracy in mathematics, he allows
> the AI to have ÃfffÃf,Ã,¢?opersonal thoughtsÃfffÃf,Ã,¢?Ãff,Ãf,Ã,. The AI
> equivalent of
> mindfulness
> (straight out of Marcus Aurelius). He gives the impression that he feeds
> back
> the AI's response to the AI, so the AI can have a think about what it is
> thinking. A bit like briefing a fish on the concept of water.
> 
> This supports my pitch for engineers to be taught philosophy. The concepts
> of virtue and evil were defined simply 2000 years ago. These definitions
> have
> stood outside of time and can be implemented with appropriate pattern
> matching.
> If Marcus Aurelius was reincarnated, he'd be running international
> seminars
> on mindfulness. The foundation ideas have not changed a wit.
> 
> Cheers
> Les
> 
> C2 - Confidential
> 
> Answer to the first question. Once the NN is trained, it will always
> give the
> same output for the same input. It is stupid software logic.
> 
> Answer to the second question. The current safety paradigm is based on
> perception-interpretation-decision-action loops based on the fact that:
> 1 Either, all states of the system, leading to all identified losses,
> can be
> exhaustively defined and their probability of occurrences remains, by
> concept and construction of the system, below a tolerable level, enabling
> fully autonomous functions (e.g. emergency shutdown functions in process
> industries),
> 
> 2 Or, there is a human in control of the system with the ability of
> arbitrating the decisions, meaning the ability to at least perceive and
> interpret.
> 
> AI deceives point 1 as it impossible, by construction, to identify all
> dangerous situations (the training dataset is incomplete and whatever its
> size, the operational domain being infinite (or close to infinite from a
> practical
> point of view), training dataset size divided by infinite = 0.
> 
> To deceive the second assumption, just replace a car windscreen by a
> video
> screen. No need of AI.
> 
> AttentionÃffff,Ãfff, : due to increased cybersecurity screening I may
> receive
> external
> emails 30 minutes after their emission.
> RESTRICTED
> 
> Bertrand RICQUE
> Safety and security standardisation Expert Program Support Manager
> Optronics and Defence Division/Customer Support Safran Electronics &
> Defense
> 
> P +33 (0)1 58 11 96 82   M +33 (0)6 87 47 84 64
> bertrand.ricque at safrangroup.com
> 102 Avenue de Paris
> 91300 MASSY FRANCE
> https://urldefense.com/v3/__http://www.safran-electronics-
> defense.com_
> _;!!Dl6pPzL6!fVuNoqxkdCCqescc5hd8-9ke7-
> o3uVuZEOBtugdgtfrHUBcO3T2a6LyI_
> lZBwk7CZBJRmOhVQnZHXCdvIB_h5MY$
> 
> -----Message d'origine-----
> DeÃffff,Ãfff, : Les Chambers <les at chambers.com.au>
> EnvoyÃfffffÃfff,Ãff,Ãf,Ã,©Ãffff,Ãfff, :
> jeudi 22
> juin
> 2023 23:46 ÃfffffÃfffÃf,Ã,¢,Ãff,Ãf,Ã,¬Ãffff,Ãfff, : RICQUE Bertrand (SAFRAN
> ELECTRONICS &
> DEFENSE)
> <bertrand.ricque at safrangroup.com>; koopman.cmu at gmail.com;
> systemsafety at lists.techfak.uni-bielefeld.de
> ObjetÃffff,Ãfff, : RE: [SystemSafety] AI and the virtuous test Oracle
> 
> CAUTION:  This message originated from an outside organization. In
> case of suspicion, click on "Report to SAFRAN Security" from the
> Outlook
> ribbon.
> 
> RE your comment, ÃffffÃff,Ãf,Ã,¢ÃfffÃf,Ã,¢,Ãff,Ãf,Ã,¬Ãfff."A neural 
network is
> software
> logic.ÃffffÃff,Ãf,Ã,¢ÃfffÃf,Ã,¢,Ãff,Ãf,Ã,¬Ãfff,Ãff,Ãf,Ã,
> 
> A neural Network is a component of an AI agent. An AI agent is a
> suite of general purpose software tools. The agent is configured from a
> dataset.
> 
> ChatGPT goes on (see dialog below):
> ÃffffÃff,Ãf,Ã,¢ÃfffÃf,Ã,¢,Ãff,Ãf,Ã,¬Ãfff."Synthesizing an AI
> agent
> involves training the machine learning model on a large
> dataset.ÃffffÃff,Ãf,Ã,¢ÃfffÃf,Ã,¢,Ãff,Ãf,Ã,¬Ãfff,Ãff,Ãf,Ã,
> 
> This indicates that the training dataset determines the
> agentÃffffÃff,Ãf,Ã,¢ÃfffÃf,Ã,¢,Ãff,Ãf,Ã,¬ÃfffÃf,Ã,¢"Ãff,Ãf,Ã,¢s
> behaviour (transfer function - inputs to outputs). In the case of a
> Tesla motor
> vehicle,
> the inputs are camera sensors and the outputs will soon be controls
> to the accelerator, brake and steering.
> 
> RE your comment, ÃffffÃff,Ãf,Ã,¢ÃfffÃf,Ã,¢,Ãff,Ãf,Ã,¬Ãfff."The problem is 
not AI. It
> is the
> exhaustive
> deterministic
> relationship between the undesired losses and the system states.
> Anything breaking this, breaks the current safety paradigm. And AI
> is not the single thing to achieve
> this.ÃffffÃff,Ãf,Ã,¢ÃfffÃf,Ã,¢,Ãff,Ãf,Ã,¬Ãfff,Ãff,Ãf,Ã,
> 
> Could you please expand on this idea. Possibly using an example.
> 
> How would your thought apply here: I am a driver at the wheel of a
> Tesla travelling at 110 km/h on the freeway. I have switched the
> vehicle to full
> self-
> driving mode. I am fast asleep. The FSD agent detects a false
> positive ÃffffÃff,Ãf,Ã,¢ÃfffÃf,Ã,¢,Ãff,Ãf,Ã,¬ÃfffÃf,Ã,¢?o
> a lady
> with a pram. It jams on the brakes initiating a rear-end collision
> with a following semi.
> Q: In this case how is it that, 
ÃffffÃff,Ãf,Ã,¢ÃfffÃf,Ã,¢,Ãff,Ãf,Ã,¬Ãfff."The problem
> is not
> AIÃffffÃff,Ãf,Ã,¢ÃfffÃf,Ã,¢,Ãff,Ãf,Ã,¬Ãfff,Ãff,Ãf,Ã,
> 
> Les
> 
> ÃffffÃff,Ãf,Ã,¢ÃfffÃf,Ã,¢,Ãff,Ãf,Ã,¬ÃfffÃf,Ã,¢?
> Ãff,Ãf,Ã,ÃffffÃff,Ãf,Ã,¢ÃfffÃf,Ã,¢,Ãff,Ãf,Ã,¬ÃfffÃf,Ã,¢?
> Ãff,Ãf,Ã,ÃffffÃff,Ãf,Ã,¢ÃfffÃf,Ã,¢,Ãff,Ãf,Ã,¬ÃfffÃf,Ã,¢?
> Ãff,Ãf,Ã,ÃffffÃff,Ãf,Ã,¢ÃfffÃf,Ã,¢,Ãff,Ãf,Ã,¬ÃfffÃf,Ã,¢?
> Ãff,Ãf,Ã,ÃffffÃff,Ãf,Ã,¢ÃfffÃf,Ã,¢,Ãff,Ãf,Ã,¬ÃfffÃf,Ã,¢?Ãff,Ãf,Ã,
> ChatGPT Dialog
> ÃffffÃff,Ãf,Ã,¢ÃfffÃf,Ã,¢,Ãff,Ãf,Ã,¬ÃfffÃf,Ã,¢?
> Ãff,Ãf,Ã,ÃffffÃff,Ãf,Ã,¢ÃfffÃf,Ã,¢,Ãff,Ãf,Ã,¬ÃfffÃf,Ã,¢?
> Ãff,Ãf,Ã,ÃffffÃff,Ãf,Ã,¢ÃfffÃf,Ã,¢,Ãff,Ãf,Ã,¬ÃfffÃf,Ã,¢?
> Ãff,Ãf,Ã,ÃffffÃff,Ãf,Ã,¢ÃfffÃf,Ã,¢,Ãff,Ãf,Ã,¬ÃfffÃf,Ã,¢?
> Ãff,Ãf,Ã,ÃffffÃff,Ãf,Ã,¢ÃfffÃf,Ã,¢,Ãff,Ãf,Ã,¬ÃfffÃf,Ã,¢?
> Ãff,Ãf,Ã,ÃffffÃff,Ãf,Ã,¢ÃfffÃf,Ã,¢,Ãff,Ãf,Ã,¬ÃfffÃf,Ã,¢?Ãff,Ãf,Ã,
> 
> ChatGPT prompt:
> ÃffffÃff,Ãf,Ã,¢ÃfffÃf,Ã,¢,Ãff,Ãf,Ã,¬Ãfff."Describe an AI agent from a 
technical
> perspective. What
> are its
> components
> and
> how is it synthesized?
ÃffffÃff,Ãf,Ã,¢ÃfffÃf,Ã,¢,Ãff,Ãf,Ã,¬Ãfff,Ãff,Ãf,Ã,
> 
> Response:
> An AI agent, from a technical perspective, typically consists of
> several components that work together to perform specific tasks.
> Here are the key components and how they are synthesized:
> 
> 1. Data: The foundation of any AI agent is data. It can be in
> various forms such as text, images, audio, or structured data. This
> data is used for
> training
> and providing context to the AI model.
> 
> 2. Machine Learning Model: The AI agent utilizes a machine learning
> model
> to
> process and generate responses. These models are trained on vast
> amounts of data and learn patterns to make predictions or generate
> outputs.
> 
> 3. Natural Language Processing (NLP): NLP is a crucial component for
> understanding and processing human language. It involves tasks like
> language translation, sentiment analysis, intent recognition, and
> entity extraction.
> NLP
> enables the AI agent to comprehend user inputs and generate relevant
> responses.
> 
> 4. Knowledge Representation: Knowledge representation refers to how
> information is stored and organized within the AI agent. It can
> employ various
> techniques
> such as semantic networks, ontologies, or graph databases to store
> facts
> and
> relationships. This allows the agent to access and utilize knowledge
> when responding to user queries.
> 
> 5. Dialogue Management: Dialogue management focuses on handling the
> flow of conversation. It involves tracking the dialogue context,
> managing user requests and responses, and generating meaningful and
> coherent replies.
> Reinforcement
> learning techniques can be applied to improve dialogue management.
> 
> 6. Integration and APIs: AI agents can integrate with external
> systems and services using APIs (Application Programming
> Interfaces). This allows them
> to
> access relevant data sources, databases, or external functionalities
> to provide more comprehensive and accurate responses.
> 
> 7. User Interface: The AI agent needs a user interface to interact
> with
> users
> effectively. It can be a chat interface, voice-based interface, or a
> combination of both. The UI should provide a seamless experience for
> users to interact with the agent.
> 
> Synthesizing an AI agent involves training the machine learning
> model on a large dataset. This process includes steps like
> preprocessing the data, designing the model architecture, selecting
> appropriate algorithms,
> training
> the model using optimization techniques, and fine-tuning it for
> specific
> tasks.
> The trained model is then integrated with other components, such as
> NLP, knowledge representation, and dialogue management, to create a
> functioning AI agent. Regular updates, feedback loops, and
> continuous improvement are essential to enhance the agent's
> performance over time.
> 
> --------END Chat ----------
> 
> C2 - Confidential
> 
> The problem is not AI. It is the exhaustive deterministic
> relationship
> between the undesired losses and the system states. Anything
> breaking this, breaks the current safety paradigm. And AI is not the
> single thing to
> achieve
> this.
> 
> AttentionÃfffffÃfffÃf,Ã,¢?sÃffff,Ãfff, : due to increased cybersecurity
> screening I may
> receive
> external
> emails 30 minutes after their emission.
> RESTRICTED
> 
> Bertrand RICQUE
> Safety and security standardisation Expert Program Support Manager
> Optronics and Defence Division/Customer Support Safran Electronics
> & Defense
> 
> P +33 (0)1 58 11 96 82   M +33 (0)6 87 47 84 64
> bertrand.ricque at safrangroup.com
> 102 Avenue de Paris
> 91300 MASSY FRANCE
> https://urldefense.com/v3/__http://www.safran-electronics-
> defense.com__;!!Dl6pPzL6!bg5nEqH6ID136htdDa-
> 
> DgyRz2IZw9arqA9HPKC3p01ZYHoQhzebcyOn5xrgApDNO52A_sLRMk2YgX63j
> n5fk7M4$
> 
> -----Message d'origine-----
> DeÃfffffÃfffÃf,Ã,¢?sÃffff,Ãfff, : systemsafety <systemsafety-
> bounces at lists.techfak.uni-
> bielefeld.de>
> De la part de Phil Koopman
> EnvoyÃfffffÃfff?'Ãffff,Ãfff,Ãff,Ãf,Ã,©ÃfffffÃfffÃf,Ã,¢?sÃffff,Ãfff, : jeudi 
22 juin
> 2023 03:32
> ÃfffffÃfff?'ÃffffÃff,Ãf,Ã,¢ÃfffÃf,Ã,¢?sÃfff,Ãff,Ãf,Ã,¬ÃfffffÃfffÃf,Ã,¢?
sÃffff,Ãfff, :
> les at chambers.com.au; systemsafety at lists.techfak.uni-
> bielefeld.de
> ObjetÃfffffÃfffÃf,Ã,¢?sÃffff,Ãfff, : Re: [SystemSafety] AI and the virtuous 
test
> Oracle
> 
> CAUTION:  This message originated from an outside organization.
> In case
> of
> suspicion, click on "Report to SAFRAN Security" from the Outlook
> ribbon.
> 
> Les,
> 
> Since you welcome riffs, I have something that is not as all-
> encompassing,
> but might have more immediate application.
> 
> I propose that to the degree that "AI" technology is deployed in
> a way
> that
> supplants practical human judgement, the manufacturer of that
> system
> (in
> some cases just the AI part if it is an add-on component) should
> be
> held
> accountable for any action (or inaction) that, if associated
> with the
> human
> that was supplanted, would have constituted negligence.ÃfffffÃfffÃf,Ã,¢?
> sÃffff,Ãfff,
> This
> should
> include
> situations in which a human is put in an untenable situation of
> supervising
> an
> AI in a way that puts unreasonable demands upon them, amounting
> to a "moral crumple zone"
> approach
> 
> (https://urldefense.com/v3/__https://papers.ssrn.com/sol3/papers.cfm
> ?ab
> 
> stract_id=2757236__;!!Dl6pPzL6!dV6V79CEWJVLcdXXS5n2wYWdaCGJCzdLlz4
> gg9Cz063kcikC8CIr0YMf2lF9o5xNrnA0Av-DS0QOuOFaUivQZX7h$ ).
> Liability/negligence if an AI is in substantive control of such
> a
> situation
> should
> attach to the manufacturer.
> 
> This leads to a more narrow oracle, but perhaps still useful,
> than you propose. If a loss event is caused by a lack of
> "reasonable" behavior
> by an
> AI,
> the manufacturer is on the hook for negligence, and the
> AI/manufacturer owes a duty of care the same as the human who
> was supplanted would
> have
> owed to whoever/whatever might be affected by that negligence.
> It has
> the
> advantage of reusing existing definitions of "reasonable person"
> that
> have
> been hammered out over decades of law. (To be sure that is not
> in the
> form
> of an engineering specification, but case law has a pretty
> robust set
> of
> precedents, such as crashing into something after your properly
> functioning
> vehicle ran a red light is likely to lead to the driver being
> found
> negligent.)
> 
> This does not require the AI to behave the same as people, and
> is not a
> full
> recipe for "safe" AI. But it puts a floor on things in a way
> that is
> readily
> actionable using existing legal mechanisms and theories. If a
> reasonable
> person would have avoided a harm, any AI that fails to avoid the
> harm
> would
> be negligent.
> 
> I've worked with a lawyer to propose this approach for automated
> vehicles,
> and it is starting to get some traction. What I write in this
> post
> (above)
> is a
> generalization of the concept beyond the narrow automated
> vehicle application.
> Details here:
> 
> https://urldefense.com/v3/__https://safeautonomy.blogspot.com/2023/0
> 5
> /a-liability-approach-for-
> 
> automated.html__;!!Dl6pPzL6!dV6V79CEWJVLcdXXS5n2wYWdaCGJCzdLlz4gg
> 9Cz063kcikC8CIr0YMf2lF9o5xNrnA0Av-DS0QOuOFaUh295b5I$
> 
> -- Phil
> 
> On 6/21/2023 7:14 PM, Les Chambers wrote:
> Hi All
> 
> I find myself reflecting on what will become of us.
> As systems engineering best practice is overrun by AI.
> 
> Practitioners report that neural networks are eating code.
> Example 1: The vector field surrounding a Tesla motor vehicle
> is an output of a neural network, not the result of software
> logic. Soon
> the
> neural net - not code - will generate controls. The size of
> the code base is reducing.  (Elon
> Musk)
> Example 2: the ChatGPT transformer code base is only 2000 LOC
> (Mo Gawdat
> https://urldefense.com/v3/__https://youtu.be/bk-
> nQ7HF6k4__;!!Dl6pPzL6!
> 
> dV6V79CEWJVLcdXXS5n2wYWdaCGJCzdLlz4gg9Cz063kcikC8CIr0YMf2lF9o5xN
> rnA0Av
> -DS0QOuOFaUpIawiVG$ )
> 
> The intelligence resides in terabytes of data, perceptrons and
> millions of weighting parameters. All are gathered by
> automated
> means.
> Not subject to human review.
> 
> Ergo what will become of our trusty barriers to dangerous
> failure:
> 1. Safety functions - gone
> 2. Verification - gone
> 3. Code reviews - gone
> 4. Validation - How?
> 
> On validation, may I suggest the moral AI. A test oracle built
> on a virtuous dataset, capable of interrogating the target
> system to determine virtue. Test outcomes will morph from
> pass/failure to
> moral/immoral.
> 
> Credible industry players have predicted that soon we will
> have AIs orders of magnitude smarter than us. Especially when
> they start
> talking
> to
> each other.
> The bandwidth will be eye-watering - the increase in
> intelligence,
> vertical.
> 
> New barriers are required. Time to develop an AI that is on
> our side
> ÃfffffÃfff,Ãff,Ãf,Ã,¢ÃffffÃff,Ãf,Ã,¢ÃfffÃf,Ã,¢?
> sÃfff,Ãff,Ãf,Ã,¬ÃffffÃff,Ãf,Ã,¢ÃfffÃf,Ã,¢,Ãff,Ãf,Ã,¬Ãfff."
> the side of ethics and the moral life. An adult in the room if
> you like. We should birth this creature now and raise it as good
> parents.
> 
> Let us not panic. May I put the proposition: virtue, like
> creativity, can be algorithmic.
> I have a sense of starting from the beginning - tabula rasa. I
> suggest
> that high-level thinking on the subject could begin with
> ChatGPT
> prompts:
> 1. What is the stoic 
philosopherÃfffffÃfff,Ãff,Ãf,Ã,¢ÃffffÃff,Ãf,Ã,¢ÃfffÃf,Ã,¢?
> sÃfff,Ãff,Ãf,Ã,¬ÃffffÃff,Ãf,Ã,¢ÃfffÃf,Ã,¢?
> zÃfff,Ãff,Ãf,Ã,¢s concept of
> virtue?
> 2. What are the elements of philosophy relevant to AI?
> 
> Let us not forget our engineering mission: Guardians of the
> divine Logos, the organizing principle of the universe,
> responsible for its creation, maintenance, and order.
> 
> Would anyone care to riff on this?
> 
> Les
> 
> --
> 
> Les Chambers
> 
> les at chambers.com.au
> systemsengineeringblog.com
> 
> +61 (0)412 648 992
> _______________________________________________
> The System Safety Mailing List
> systemsafety at TechFak.Uni-Bielefeld.DE
> Manage your subscription:
> https://urldefense.com/v3/__https://lists.techfak.uni-
> bielefeld.de/mai
> 
> lman/listinfo/systemsafety__;!!Dl6pPzL6!dV6V79CEWJVLcdXXS5n2wYWdaC
> GJCz
> dLlz4gg9Cz063kcikC8CIr0YMf2lF9o5xNrnA0Av-
> DS0QOuOFaUjXMdlNF$
> 
> --
> Prof. Phil Koopman   koopman at cmu.edu
> (he/him)
> 
> https://urldefense.com/v3/__https://users.ece.cmu.edu/*koopman/__;fg!
> 
> !Dl6pPzL6!dV6V79CEWJVLcdXXS5n2wYWdaCGJCzdLlz4gg9Cz063kcikC8CIr0YM
> f2lF9o5xNrnA0Av-DS0QOuOFaUnsUrevc$
> 
> _______________________________________________
> The System Safety Mailing List
> systemsafety at TechFak.Uni-Bielefeld.DE
> Manage your subscription:
> https://urldefense.com/v3/__https://lists.techfak.uni-
> 
> bielefeld.de/mailman/listinfo/systemsafety__;!!Dl6pPzL6!dV6V79CEWJVL
> cd
> XXS5n2wYWdaCGJCzdLlz4gg9Cz063kcikC8CIr0YMf2lF9o5xNrnA0Av-
> DS0QOuOFaUjXMdlNF$
> 
> #
> " Ce courriel et les documents qui lui sont joints peuvent
> contenir des
> informations confidentielles, ÃfffffÃfff?'Ãffff,Ãfff,Ãff,Ãf,Ã,ªtre soumis 
aux
> rÃfffffÃfff?'Ãffff,Ãfff,Ãff,Ãf,Ã,¨glementations
> relatives au
> contrÃfffffÃfff?'Ãffff,Ãfff,Ãff,Ãf,Ã,´le des exportations ou ayant un
> caractÃfffffÃfff?'Ãffff,Ãfff,Ãff,Ãf,Ã,¨re
> privÃfffffÃfff?'Ãffff,Ãfff,Ãff,Ãf,Ã,©.
> S'ils ne
> vous sont
> pas destinÃfffffÃfff?'Ãffff,Ãfff,Ãff,Ãf,Ã,©s, nous vous signalons qu'il est
> strictement
> interdit
> de les divulguer, de les reproduire ou d'en utiliser de quelque
> maniÃfffffÃfff?'Ãffff,Ãfff,Ãff,Ãf,Ã,¨re que ce
> soit
> le contenu. Toute exportation ou 
rÃfffffÃfff?'Ãffff,Ãfff,Ãff,Ãf,Ã,©exportation
> non
> autorisÃfffffÃfff?'Ãffff,Ãfff,Ãff,Ãf,Ã,©e
> est interdite Si ce message vous a
> ÃfffffÃfff?'Ãffff,Ãfff,Ãff,Ãf,Ã,©tÃfffffÃfff?'Ãffff,Ãfff,Ãff,Ãf,Ã,©
> transmis par
> erreur,
> merci d'en informer l'expÃfffffÃfff?'Ãffff,Ãfff,Ãff,Ãf,Ã,©diteur et de 
supprimer
> immÃfffffÃfff?'Ãffff,Ãfff,Ãff,Ãf,Ã,©diatement
> de votre systÃfffffÃfff?'Ãffff,Ãfff,Ãff,Ãf,Ã,¨me informatique ce courriel 
ainsi
> que tous
> les
> documents qui y sont attachÃfffffÃfff?'Ãffff,Ãfff,Ãff,Ãf,Ã,©s."
> ******
> " This e-mail and any attached documents may contain confidential
> or
> proprietary information and may be subject to export control laws
> and regulations. If you are not the intended recipient, you are
> notified that
> any
> dissemination, copying of this e-mail and any attachments thereto or
> use of their contents by any means whatsoever is strictly
> prohibited. Unauthorized export or re-export is prohibited. If you
> have received this e-mail in
> error,
> please advise the sender immediately and delete this e-mail and all
> attached
> documents from your computer system."
> #
> 
> --
> 
> Les Chambers
> 
> les at chambers.com.au
> 
> +61 (0)412 648 992
> 
> #
> " Ce courriel et les documents qui lui sont joints peuvent contenir
> des
> informations confidentielles, ÃfffffÃfff,Ãff,Ãf,Ã,ªtre soumis aux
> rÃfffffÃfff,Ãff,Ãf,Ã,¨glementations
> relatives au
> contrÃfffffÃfff,Ãff,Ãf,Ã,´le des exportations ou ayant un
> caractÃfffffÃfff,Ãff,Ãf,Ã,¨re
> privÃfffffÃfff,Ãff,Ãf,Ã,©.
> S'ils ne
> vous sont
> pas destinÃfffffÃfff,Ãff,Ãf,Ã,©s, nous vous signalons qu'il est strictement
> interdit
> de les
> divulguer, de les reproduire ou d'en utiliser de quelque
> maniÃfffffÃfff,Ãff,Ãf,Ã,¨re
> que ce
> soit
> le contenu. Toute exportation ou rÃfffffÃfff,Ãff,Ãf,Ã,©exportation non
> autorisÃfffffÃfff,Ãff,Ãf,Ã,©e est
> interdite Si ce message vous a ÃfffffÃfff,Ãff,Ãf,Ã,©tÃfffffÃfff,Ãff,Ãf,Ã,© 
transmis
> par erreur,
> merci
> d'en
> informer l'expÃfffffÃfff,Ãff,Ãf,Ã,©diteur et de supprimer
> immÃfffffÃfff,Ãff,Ãf,Ã,©diatement de
> votre
> systÃfffffÃfff,Ãff,Ãf,Ã,¨me informatique ce courriel ainsi que tous les
> documents qui
> y
> sont
> attachÃfffffÃfff,Ãff,Ãf,Ã,©s."
> ******
> " This e-mail and any attached documents may contain confidential or
> proprietary information and may be subject to export control laws and
> regulations. If you are not the intended recipient, you are notified that
> any
> dissemination, copying of this e-mail and any attachments thereto or use
> of
> their contents by any means whatsoever is strictly prohibited.
> Unauthorized
> export or re-export is prohibited. If you have received this e-mail in
> error,
> please advise the sender immediately and delete this e-mail and all
> attached
> documents from your computer system."
> #
> 
> --
> 
> Les Chambers
> 
> les at chambers.com.au
> 
> +61 (0)412 648 992
> 
> #
> " Ce courriel et les documents qui lui sont joints peuvent contenir des
> informations confidentielles, ÃffffÃff,Ãf,Ã,ªtre soumis aux
> rÃffffÃff,Ãf,Ã,¨glementations
> relatives
> au
> contrÃffffÃff,Ãf,Ã,´le des exportations ou ayant un caractÃffffÃff,Ãf,Ã,¨re
> privÃffffÃff,Ãf,Ã,©.
> S'ils ne vous
> sont
> pas destinÃffffÃff,Ãf,Ã,©s, nous vous signalons qu'il est strictement 
interdit
> de
> les
> divulguer, de les reproduire ou d'en utiliser de quelque 
maniÃffffÃff,Ãf,Ã,¨re
> que
> ce
> soit
> le contenu. Toute exportation ou rÃffffÃff,Ãf,Ã,©exportation non
> autorisÃffffÃff,Ãf,Ã,©e
> est
> interdite Si
> ce message vous a ÃffffÃff,Ãf,Ã,©tÃffffÃff,Ãf,Ã,© transmis par erreur, 
merci d'en
> informer
> l'expÃffffÃff,Ãf,Ã,©diteur
> et de supprimer immÃffffÃff,Ãf,Ã,©diatement de votre systÃffffÃff,Ãf,Ã,¨me
> informatique ce
> courriel
> ainsi
> que tous les documents qui y sont attachÃffffÃff,Ãf,Ã,©s."
> ******
> " This e-mail and any attached documents may contain confidential or
> proprietary information and may be subject to export control laws and
> regulations. If you are not the intended recipient, you are notified that
> any
> dissemination, copying of this e-mail and any attachments thereto or use
> of
> their contents by any means whatsoever is strictly prohibited.
> Unauthorized
> export or re-export is prohibited. If you have received this e-mail in
> error,
> please advise the sender immediately and delete this e-mail and all
> attached
> documents from your computer system."
> #
> 
> --
> 
> Les Chambers
> 
> les at chambers.com.au
> 
> +61 (0)412 648 992
> _______________________________________________
> The System Safety Mailing List
> systemsafety at TechFak.Uni-Bielefeld.DE
> Manage your subscription: https://lists.techfak.uni-
> bielefeld.de/mailman/listinfo/systemsafety
> 
> --
> 
> Les Chambers
> 
> les at chambers.com.au
> 
> +61 (0)412 648 992
> 
> --
> 
> Les Chambers
> 
> les at chambers.com.au
> 
> +61 (0)412 648 992
> 
> --
> 
> Les Chambers
> 
> les at chambers.com.au
> 
> +61 (0)412 648 992



--

Les Chambers

les at chambers.com.au

+61 (0)412 648 992