Cunning clones
Techno Talk
Imagine that you could hold a conversation with holographic clones
of your great-grandparents, asking them questions about their
lives, loves, hopes and fears. It’s too late for us, but this may be a
possibility for our own great-grandchildren.
I
n my previous column, I introduced
a company called Proto (see their
website at protohologram.com). Its
claim to fame is the ability to generate
artificial intelligence (AI) based full-size
3D volumetric holograms of people.
They start by videoing a conversation with the person in question.
Let’s consider William Shatner—who
famously played Captain James Tiberius
Kirk in Star Trek: The Original Series
(TOS)—as an example because they’ve
already done him. The idea was to
create a hologram of Bill that could
attend shows and conferences and
answer both Shatner- and Star Trekrelated questions.
Proto’s AI used this video to create
what I think of as a ‘voice clone’. The
AI can subsequently use this clone
to say anything it likes while sounding just like Bill. Their AI also creates
what I think of as a ‘physical clone’ or
a ‘body clone’, which it uses to generate the full-size 3D hologram. This
hologram is generated on a frame-byframe basis in real-time.
Any visual aspects of the hologram—
lips, facial muscles, eye blinks, hand
and body gestures—are synchronised
with whatever it’s saying.
Who are you?
To be capable of holding a Turing test
level conversation, the AI powering the
hologram needs what I think of as a
‘knowledge clone’ of the person in question. In the case of Bill, the AI can build
this knowledge clone by interviewing
him, starting with basic questions like,
“What’s your name?”, “Where were you
born?”, “How old are you?” and “How
do you pronounce ‘sabotage’?”.
Each question can open new avenues
of conversational exploration, just like
a chat between two people.
If they are lucky enough to get to
meet Bill, many people want to pose
Star Trek-related questions. Although
Bill generally enjoys answering queries
of this nature, he sometimes expresses
frustration if he feels the questions are
repetitive or overly focused on specific
details. Unfortunately, fiddly details are
what Star Trek fans like most.
56
The folks at Proto got around this by
feeding the AI with every script from
every episode, thereby allowing the
hologram to answer any conceivable
tribble-connected question without
trouble (obviously everyone knows
Season 2, Episode 15: “The Trouble
with Tribbles”).
Max the Magnificent
In the 1974 film Young Frankenstein,
Frederick sends hunchback Igor to the
local “Brain Depository” to steal the
brain of a deceased “scientist and saint”
called Hans Delbrück. Unfortunately,
Igor accidentally drops the glass jar
containing the brain in question, so
he grabs another close by.
He tells Frederick the brain came
from someone called “Abby Normal”,
but later admits the jar was, in fact, labelled “Abnormal”.
With Proto’s holograms, it’s possible to
swap one knowledge clone (the ‘brain’)
for another. Suppose that instead of
loading our Shatner AI with Star Trek
scripts, we instead fed it with books on
infectious diseases. This would enable
it to hold forth on medical matters (just
don’t ask it about Tribbles).
transportation hubs like airports (their
human hosts always seem to be absent
when I need them the most).
Or how about having one at the end
of every aisle in a gigantic hardware
store? These AIs can see and hear you,
so you could show them something
and ask, “What’s the name of this and
where can I find one?” With its subject
matter expertise, the AI might respond,
“That’s called a ballcock valve, and
you’ll find them on the bottom shelf
halfway down aisle 10 on the right.”
A lot of people are embarrassed to
talk to doctors about certain things. But
experiments have shown that people
are much more forthcoming with AIs,
especially ones like Proto’s holograms.
And, just like a human doctor, the AI
could be watching the patient, observing things like involuntary movements
of the hands or rapid blinking or…
I can easily envisage these AI holograms in nursing facilities, especially
those dealing with people suffering
cognitive decline like dementia. Unlike
human carers, an AI wouldn’t mind
answering the same question multiple
times or listening to the same story over
and over (and over) again.
Who said that?
The Life of (a boy called) Clive
Abby Normal
The AIs powering Proto’s holograms
can typically recognise and speak over
120 languages (excluding Australian,
of course).
Thus, if a Spanish fan approached our
William hologram and asked, “¿Cuál
es el verdadero problema con los tribbles?” (ie, “What’s the real trouble with
tribbles?”), the AI could (a) respond in
Spanish complete with apical-alveolar
trills (ie, the rolling of the Rs) and (b)
the hologram’s lips and other facial
muscles would be perfectly synchronised with the words.
What’s the point?
You might be thinking, “This is all
very clever, but what’s the point?” By
this, you mean, “What could we use
these AI holograms for?” To be honest,
we are limited only by our imaginations.
Mainstream uses for this technology
might include information booths at
I miss my dad. He passed away in
January 2000. I would love to have
access to an AI hologram of him—one
that had spent weeks chatting to him
and learning all about him. I know it
wouldn’t be like really having my dad
present, but being able to hold a conversation with him (even in holographic
form) would soothe my soul, as it were.
As I mentioned a couple of columns
ago, I have almost finished writing a
book called The Life of (a boy called)
Clive. This describes my life growing
up as a kid in Sheffield in the 1960s.
Suppose we fed the contents of this
book to an AI hologram of me. The AI
could also interview me to fill in additional details. Long after I’m gone,
my great-great-grandchildren could
chat with my holographic AI to at least
get a glimpse as to what I was like. It
would be a form of immortality. It’s
something to think about…
PE
Practical Electronics | April | 2025