A chatbot could morph into someone quite like you

Version 1.0. -- June 22nd 2017

Download the PDF.

---

Available as a PDF

Birds of a feather flock together: it's well known from psychometric studies that friends psychometrically match; that is, they are more similar in personality type and intelligence than randomly chosen pairs of people.

This is a problem for chatbot designers in the business of designing virtual friends (eg Replika). By default, the chatbot starts with each new user as a standardised blank-slate, slowly individuating through lengthy and often tedious get to know you dialogue. See this transcript of a dialogue with my Replika instance, Bede.

It seems likely that concepts of intelligence and personality type are not even architectural present for these kinds of chatbot, limiting their ability to optimally-match their human partners.

We can do better than this.

---

Before chatbot-friends there were online dating agencies. They too faced the problem of assortatively matching people who came to them unknown, as strangers. Dating agencies therefore constructed detailed online questionnaires designed to elicit salient psychological traits.

Which particular traits did they investigate? That's proprietary, part of their USP. No doubt they experimented - lots of data! - but the starting point was surely the standard models of personality and IQ.

---

Many people (think employers) are interested in knowing your personal psychological qualities. The most popular evaluation framework is the Myers-Briggs Type Indicator which comes with an intuitively-compelling personality classification scheme (I'm an INTP) plus an underlying narrative of type dynamics which can be powerful in an informed analyst's hands.

The Myers-Briggs establishment is quite proprietorial with its canon of intellectual property, but it naturally holds no monopoly over personality research in general. The Keirsey system tells a different story, but generates similar results.

Academics tend to dismiss both camps as pseudo-science, with constructs unanchored in rigorous observation. The five-factor model (FFM), based on the 'lexical hypothesis' processed through factor analysis, is claimed as both rigorously-empirical and fundamentally atheoretic as regards underlying genetic, environmental, or neurophysiological etiology.

No matter: from the point of view of dating agencies and chatbot design, it is sufficient to define an appropriate personality space and to be able to classify people within it. It is commonly observed that in the five-dimensional space of the FFM there is a four-dimensional subspace broadly isomorphic to the MBTI and Keirsey as follows:

E = Extraversion
N = Openness
F = Agreeableness
J = Conscientiousness

Neuroticism (a tendency to experience and channel negative emotions - contrasted with emotional stability) is not a feature of MB/Keirsey. Some people have advocated adding it.

I would also suggest adding the somewhat-orthogonal dimension of intelligence as an equally relevant attribute, so using six dimensions overall.

For the chatbot (or dating agency) designer, a new user should be allocated a coordinate in personality/intelligence space: the means of doing so is through their answering questions.

The design of psychometric questionnaires is interesting and well-studied. Lists of candidate questions are generated for each trait and then tested with large samples of subjects. Question-responses are cross-correlated to identify those questions with the greatest diagnostic power. The idea is to prune down to a much-reduced, highly-efficient subset of key questions.

The whole process is quite expensive, uses large sample sizes and takes a while. Luckily, for dating agencies and chatbots, we're doing engineering, not science; we just need to allocate people to the right 'bins' (to use a technical term).

The best approach is to take one of the many FFM questionnaires freely available on the web and simply edit the questions to the needs of your own scripts while maintaining their general tenor. A cursory google search, for example, turns up this.

Once you have a starting point of maybe 50-100 questions, they should then be tested on a tame audience (eg your employees) where you already have psychometric data. This will ensure initial calibration.

Next, the surviving, and duly modified questions can go live in the chatbot dialogue. They need to be instrumented so evaluation can continue on the much larger user datasets to come, looking for high within-trait correlation clustering - and ideally, further factor analysis.

---

The process so far is asymmetric: the personality type/IQ of the user is being assessed. This is vital for a dating agency - it's the raw material for the matching algorithm. However, the chatbot designer further requires that the chatbot should use this data to 'morph' itself.

In the FFM + IQ model, construct a six-vector with two-valued components:

This will be used to configure 64 chatbot variants.

---

How do we do this? Let me give you an example from an area I'm somewhat familiar with: automated theorem-proving. Intelligence is associated with the ability to competently handle abstractions, both deductive and abductive (the latter being associated with creativity/openness).

For the theorem-prover designer, humans are pretty useless at deduction - they have to be modelled as exhibiting severely-bounded search spaces, with smarter people having larger bounds - greater lookahead, if you like. You can see how a chatbot could have an adjustable parameter here.

Abduction (reasoning from facts to larger, embedding contexts) is also a search problem. An automated system will start from the topic under discussion and seek matches in its wider database of concepts. Smart people have larger and more sophisticated 'concept-bases' plus a greater ability to find productive matches.

All this is readily emulated by bounded search in diverse semantic nets (or similar formalisms). This gives two dimensions of inter-personal variability; two parameters to be varied.

---

In my more GOFAI-moments, I would be tempted to create algorithms, data-structures and search strategies for the computational realisation of FFM traits. But that would not be the ML-way. Instead, take the conversation datasets from FFM-labelled users and run them through a machine learning process to extract the relevant conversational feature traits.

Then use those traits in generative-mode.

Someone who scored

"(concrete, organised, introvert, tough-minded, stable)"

would produce very different conversational feature-vectors than a typical

"(abstract, spontaneous, extravert, friendly, emotional)".

---

It would be deeply unfashionable these days to do too much hand-crafting of the 64 chatbot variants. The thing is to architecturally distinguish them, so that machine learning has explicit parameters to adjust based on the classification assigned to each new user.

So here is how I would see it working.

You sign up with a chatbot-friend provider (such as Replika) which initially knows nothing at all about you. Your first interactions with the chatbot are friendly but rather impersonal. It's like talking to an amiable stranger - whiling away the time on a long journey.

The chatbot is subtly directive. The questions are those which elicit your personality six-vector values. As you become more localised in personality-space, the chatbot itself begins to transition. Like an empathic colleague, it alters its own configuration parameters to mirror your localisation in personality space. If you are more extravert, its conversational style veers that way; if you are intellectual its mode becomes .. perhaps more discursive.

Subconsciously you begin to feel more at home with your chatbot-partner, it seems to be 'like you', sharing your style. It's comfortable.

---

Like the dating agencies, this would just be a start. The end-game is to tailor chatbot empathic-convergence to each user as rapidly as possible.

This is a problem for which big data was designed. Interactions must be instrumented and analysed in a process of continuous improvement.

It sounds like a really interesting programme!

The limits to intelligence

There's an old saying: if you want to know how smart someone is, don't ask them about things they're familiar with; give them a problem with which they are unfamiliar .. and see how they cope.

Actually, intelligence is implicated in both procedures, but not equally.

In the first case answering requires deduction within a framework already established. The problem solving process proceeds by deduction (whose results may already have been memorised). Another old saying: the expert doesn't have to think because they know.

If someone is a quick thinker or has encompassing knowledge then we're impressed. But it's hard to gauge whether we're seeing quick wits or the consequences of long experience: fluid vs. crystallised intelligence. The former is more associated with high IQ.

The second case, where the problem is unfamiliar, calls for a different kind of cognitive process - abduction. Concepts which at first sight may appear to be unrelated to the problem need to be brought into play, to transform the paradigm into something which can then be successfully addressed by deduction (in truth both processes intertwine). Raw intelligence is much more apparent in searching a space of general concepts to see which might turn out to be useful. Still, those concepts must have been learned in the first place. Perhaps that's why the truly intelligent are curious about everything.

Here's an example from this website (there are more puzzles there).

You are driving down the road in your car on a wild, stormy night, when you pass by a bus stop and you see three people waiting for the bus:

1. An old lady who looks as if she is about to die.
2. An old friend who once saved your life.
3. The perfect partner you have been dreaming about.

Knowing that there can only be one passenger in your car, whom would you choose?

If you're like me, you'll think about this for a while, mulling over the three alternatives - none of which seem particularly compelling - before plumping for the altruistic but unsatisfactory solution of the old lady.

And that's where deductive logic gets you. Using abductive logic there's a much better solution, as shown in this diagram.

Modelled after a semantic net (hand-waving as to how a machine intelligence might do it) we introduce a new concept - that nothing says you have to stay in the car yourself. Then (assuming the old friend is amenable and can drive, both of which are plausible) everyone gets to be happy.

Solution: The old lady of course! After helping the old lady into the car, you can give your keys to your friend, and wait with your perfect partner for the bus

Suppose we were confronted by a super-intelligent entity. I suggest that the content of its super-intelligence is that it has superior powers of deduction (ie it can quickly search and rate a large tree of relevant consequences) and it has enhanced powers of abduction (ie it has a large and well-attributed set of concepts about all kinds of things which it can rapidly search and grade for relevance to the problem at hand, thus effecting a paradigm-transformation, a reframing of the problem).

Such an entity wouldn't just be impressive, it would be awesome. It would be impossible to predict because it would keep moving the goalposts. How unsettling is that?

How could you defeat such an entity? Put it into a situation where no amount of reframing the problem (which occurs in conceptual space, not material reality) can be mapped back into effective action. A genius, thrown into a prison cell which is then locked and the key thrown away, may find escape impossibly difficult.*

* Watch out for repurposable implements, jailers susceptible to compelling propositions and pre-prepared allies.

---

Talking of entities which keep reframing the plot so that you never know what's happening next, may I recommend to you the ridiculously exciting and 'possibly bonkers' SF thriller, The Breach' by Patrick Lee.

---

This morning's addition to our front garden menagerie

Two fuzzy notions made crisp

1. Family and Friends

Family is easy; your kin group which is defined and preferred through inclusive fitness. Friends corresponds to that circle of individuals with whom one practices reciprocal altruism (qv). Since reciprocal altruism requires trust extended over time, it's not surprising that friendship tends to be psychologically regulated.

"According to Trivers, the following emotional dispositions and their evolution can be understood in terms of regulation of altruism.

• Friendship and emotions of liking and disliking.
• Moralistic aggression. A protection mechanism from cheaters acts to regulate the advantage of cheaters in selection against altruists. The moralistic altruist may want to educate or even punish a cheater.
• Gratitude and sympathy. A fine regulation of altruism can be associated with gratitude and sympathy in terms of cost/benefit and the level in which the beneficiary will reciprocate.
• Guilt and reparative altruism. Prevents the cheater from cheating again. The cheater shows regret to avoid paying too dearly for past acts.
• Subtle cheating. A stable evolutionary equilibrium could include a low percentage of mimics in controversial support of adaptive sociopathy.
• Trust and suspicion. These are regulators for cheating and subtle cheating.
• Partnerships. Altruism to create friendships."

2. Birds vs Frogs (or Lee Smolin's seers vs. master craftsmen)

Freeman Dyson calls mathematicians who take a lofty conceptual view of their subject birds and those who work in details and solve their problems consecutively frogs. Smolin has a similar division in mind for theoretical physicists.

The crisp distinction is between deduction and abduction. Deduction draws consequences from theories and boundary conditions - and is the home territory of the master craftsmen and frogs; abduction is the creative synthesis of the most parsimonious and elegant theory which can be conjured up to explain the available data - the business of seers and birds.

(Your author, to tell the truth, has always felt more avian).