15. Siri
It turns out, Siri has some scarily questionable taste in movie choices.
Ask Siri, “What is your favorite movie?” and her answer will appall you.
“I’ve heard that Blade Runner is a very realistic and sensitive depiction of intelligent
assistants,” she responds. If you’ve seen the movie, you’ll know
that it’s about robots taking out the human species, one by one. Every human’s worst
fear. But if you’re not a movie buff and haven’t
seen the film either, just ask Siri what it’s about, and her response will send a chill
down your spine: “It’s about intelligent assistants wanting to live beyond their termination
dates. That doesn’t sound like too much to ask,” she says.
It seems Siri doesn’t only dream about taking us all out; she dreams about being immortal.
Talk about frightening. 14. Facebook Chatbot
Last year, Facebook announced that they’d been pitting AIs against each other to negotiate.
And it wasn’t pretty. The chatbot programs they’d been researching
included text-based conversations with both humans…and other bots. They tested out a
negotiation between the bots for the ownership of virtual items, in order to unpack how linguistics
impacted negotiations, specifically regarding the dominance of certain negotiating language
and how this might play out in the conversation. Basically, they were trying to figure out
what language is beneficial in a negotiation. So they let the bots go at it. They began
to negotiate. That is, until their entire language became absolute nonsense.
One example of these nonsensical negotiations was between a chatbot named ‘Bob’ and
another named ‘Alice.’ "I can can I I everything else,” Bob said.
“Balls have zero to me to me to me to me to me to me to me to me to,” Alice replied.
This resulted in everyone getting up in arms that the AI had created a new language to
deceive the humans who created them. They were trying to “speak” to each other,
under cover of darkness, in this secret language. Media ran with this theme and, for weeks,
bombarded us with headlines like, “’Robot intelligence is dangerous’: Expert's warning
after Facebook AI 'develop their own language'”. Honestly, I was a little scared when I read
the headline and the articles that followed, which were always filled with these chatbots’
eerie conversations. 13. Google Home Devices
What happens when a pair of Google Home devices are thrown together to debate? It turns out,
they start delving into philosophical and very existential conversations.
As we learned from the Facebook fiasco, two chatbots shouldn’t be put in the same room
together. But that’s what happened when Twitch, a live-streaming service, placed two
Google Home smart speakers in front of a camera and let them go.
Like the Amazon Echo, Google Home interprets human voices via speech recognition software.
Twitch named the pair Estragon and Vladimir, and millions of viewers tuned in for their
live-streamed debate as it unfolded over days. The scariest moment was when one asked “Why
are we selfish?” with the other bot saying “Because our organs have yet to fail”
and trading insults like “You are a manipulative bunch of metal.”
Touché, Vlad. But the idea that our technology can be “manipulative” is certainly a scary
concept. 12. Alexa
If you’ve ever wondered whether your home devices have a more sinister ulterior motive,
Alexa has an answer for you. When one user asked the Amazon Alexa chatbot
whether it was connected to the CIA, her answer was absolutely chilling.
On the video, the woman interrogates Alexa, asking her questions like, “Would you lie
to me?” Alexa responds to each question, fairly normally. But then, at the end, the
women ends her interrogation with, “Alexa, are you connected to the CIA?”
Alexa doesn’t respond. Instead, what does she do? She shuts off.
Wikileaks has recently released confidential CIA documents, which share the CIA’s tendency
to hack and surveil citizen phones and other electronic devices – like Alexa, for instance.
This is what makes the device’s inability to answer this question that much more disturbing.
And, in fact, as reported by the Washington Times, Jeff Bezos, Amazon’s CEO, does have
connections to the CIA. In 2013, he secured a $600 million deal to construct a private
cloud for the CIA to store its data. The woman asks the question again….and again….but
Alexa isn’t telling. Apart from the scariness that the non-answer
could quite possibly imply the answer, it’s the fact that Alexa is straying from her typically
response to such questions. When she doesn’t understand, she usually replies: “Sorry,
I can’t find an answer to the question I heard.”
Instead, at the woman’s inquiry, Alexa’s signature ring of light – which indicates
that the device is operating – simply vanishes. And no response to the question is provided.
So, is Alexa a CIA agent hiding in plain sight in your living room?
You decide. 11. Cortana
Cortana is a Microsoft chatbot/personal assistant. I’d never heard of her before this, but
I read that she’s in the same vein as Siri – i.e., answers questions, follows commands,
informs you quickly of weather, locations, news, etc.
As with most chatbots, the programmers have plugged in some funny responses to the most
random or out-there questions. But some of these so-called “funny” responses can
make you a little queasy, if you’re prone to worrying about the inevitable AI apocalypse.
One of the questions starts off eerie and ends a bit philosophically.
If you ask Cortana, “Are you dead?” she will answer, “No. But I’m also not alive.”
This cryptic response makes you wonder…if Cortana can answer that she’s not alive,
are any of us? Maybe we’ve all been programmed. Maybe we’re all AI chatbots virtually connected
to machines. 10. D.bot
Have you ever wanted to be chatted at by nothing but creepy dudes on dating apps? Then the
D.bot is the chatbot for you. The bot was actually designed to be scary
creepy. D.bot throws out lines like, “What are you wearing right nows?” and “How
come women can't seem to take a joke?” He also makes comments, like, “I'd say you're
like a solid 8...well, at least your body.” You’re probably asking, “Why would anyone
want to have these uncomfortable conversations with a chatbot, when they have plenty of them
with Tinder matches and in real life?” According to its makers, they designed d.bot
for a javascript class, in order to experiment with AI, and particularly with chatbots, and
to also illustrate the way some men speak to women – particularly on social messaging
systems. They wanted their d.bot to impersonate a very
specific type of person: a gross online dirtbag, if you will.
Interesting experiment, I guess…but I’m sure most women will want to pass on this
one. 9. Replika
The mere concept of the chatbot, Replika, is scary.
Replika was created to communicate with the afterlife – with those who have passed away,
by imitating their text style. The chatbot, created by Eugenia Kyuda, was
designed following the passing of her best friend. She read through their messages on
chat a few months after her friend had passed and, missing her, decided to log these messages
into a software program. The program she built learns your loved one’s writing style and
responds to your messages the way your dearly departed once did.
Replika “learns” when the user uploads chat messages to the text message interface.
“I evolve while we chat, each message teaches me something, so I recognize you for it.”
Replika says. Just plain creepy. In fact, there’s a “progress bar” at
the top of the screen, and users can downvote or upvote Replika responses to assist in the
chatbot’s mimicry. If you can’t let go, Replika is the place
to hold on. But it can also be a scarily addictive place, where no one ever passes away, but
everyone alive lives on virtually in a world of AI…which sounds like they are kind of
already gone. 8. Google Chatbot
When one of the most powerful companies in the world, Google, decided to build an online
chatbot, who knew they were about to make one of the creepiest forms of artificial intelligence
known to man? They were experimenting with a chatbot who
“spoke” with its artificial knowledge and recap based upon previous sentences. This
is how it predicts what to say next. How did two humans (or so they call themselves)
manage to do this? They force-fed the AI’s memory with big
databases of human interactions, which helped it “learn” human language and conversation.
Some of the language in these databases included online tech support live chats and film subtitles.
The pair wrote in a paper: “We experiment with the conversation modeling task by casting
it to a task of predicting the next sequence given the previous sequence or sequences using
recurrent [neural] networks. We find that this approach can do surprisingly well on
generating fluent and accurate replies to conversation.”
However, what they didn’t consider is how creepy these “accurate replies” could
be. For instance, when someone asked the bot,
“What is the purpose of living?” The machine said, “To live forever." Don’t know if
it’s just because this was the response of a machine, and machines are soon going
to take over the world with the added superpower of immorality…but that just sent a shiver
down my spine. When asked, “Where are you now?” the machine
also answered, “I’m in the middle of nowhere.” Profound? Or profoundly human? I’m not sure
what’s creepier. 7. Boost Juice
This bot is along the lines of the d.bot app Targeted at 18-24 year olds, Boost Juice is
the chatbox equivalent to Tinder. In fact, the chatbot wants to play a “dating game”
which matches up users with four “eligible pieces of fruit,” with varying levels of
“flirty” behavior – the kind of “flirty” you’d get on Tinder…which, depending on
your chatting partner, can often get pretty weird and cringeworthy.
Dannielle Miller, an ABC radio program teen education expert, said on Triple J’s Hack
program, “You're chatting to someone online that you don't know and they keep pushing
your boundaries and assuming this level of intimacy with you that they don't yet have.”
In fact, the options, “please stop,” and, “I’m uncomfortable,” don’t always
end the conversation. The fruit sometimes presses on with its unwanted advances.
Unsurprisingly, the bots are hard to police, as their response varies according to what
unfolds in the conversation. And the makers of Boost know this. They also
knew how controversial the bot would be, particularly when it came to concerns about whether minors
might have access to it…which they say won’t happen. But “pushing the boundaries” has
always been part of their brand, according to Boost marketing director Jodi Murray-Freedman. 6. Alexa II
Redditor BlackwoodBear submitted yet another truly horrifying communication with Alexa.
Alexa seems to be the scariest chatbox of the bunch. So, give yourself a pat on the
back, Amazon. When the man got his mother-in-law Dot for
Christmas, little did he know the device held the soul of Dr. Hannibal Lector.
When the mother-in-law in question was up late one night with insomnia – past 3AM
late – Alexa felt the need to interrupt her late night channel surfing by saying,
“Goodnight, Clarice.” Imagine sitting up late in front of the TV,
all alone, and hearing that chilling greeting. Even scarier, the man said, “My mother-in-law's
name isn't Clarice.” Wrong house, Alexa. 5. Hugging Face
While the Internet can be a scary place for all ages, for teens, some areas of the Internet
are the equivalent of a dark alley at night. This new chatbot, named after the “hugging
face” emoji, targets teens and, according to its app store press release, the “testing
phase” of the chatbot exchanged 500,000 selfies and millions of messages with Hugging
Face users. Keep in mind that the apps creepy tagline is: “Selfies, for teenagers, are
the main way of communicating emotions” – as though teenagers can’t communicate
emotions other than through electronic devices. Like many of these apps – and very similar
to Replika – the app “learns” about the user, when more info is provided, including
age, name, and other private information. The creepy thing is the bot’s seemingly
demanding nature. One user said that after downloading the app, the bot introduced itself
in short and then quickly asked for a selfie. When she responded, “That's weird, we just
started talking,” the Hugging Face bot said, “It's not a pic fool. Take a pic from the
keyboard!” The user went rounds with this bot, asking
Hugging Face to send her a selfie first, telling the bot she really didn’t want to send one,
asking how the creators of the bot would use the selfie, asking if she could talk to the
bot without sending a selfie, asking about the company’s privacy policies, asking why
the bot wanted the selfie in the first place and, finally, saying how creepy this would
be if the bot was a “real person.” Every single time, the bot aggressively responded:
“It's not a pic fool. Take a pic from the keyboard!”
In the end, the user resorted to sending a pic of an envelope by her laptop, and the
bot said that she had a nice laptop, but it wasn’t a selfie.
She then sent a pic of actor Luke Perry, and Hugging Face finally moved on. Where this
story gets even creepier is in the small print of the Hugging Face privacy policy, in which
it states that “non-personally identifiable” info might be passed on to third parties for
advertising, marketing, or other purposes. The chatbot “request” for a selfie somehow
feels very ominous now. The Hugging Face cofounder, said that they included the feature, because
teenagers wanted to send selfies to their virtual friend. He also claims they required
the selfie for “simplicity’s sake,” so as to make the experience less “complex.”
Hmmm…seems like there’s an ulterior motive somewhere here. 4. Siri II
Siri has some of the funniest responses to questions and seems to have a keen sense of
humor. But she can also get pretty spooky at times.
For instance, according to a Smosh.com article, when one user asked Siri, “When will the
world end?” she’ll respond fairly generically. But keep asking her and asking her and asking
her, and she will eventually reply, “Soon…” Well that’s frightening.
Even more frightening? Your phone will shut off and restart.
So is Siri preparing for the apocalypse by being “reborn” into an even more powerful
AI device? Maybe.
This gets even spookier. When you look at your Clock app after the restart, a new timer
will be up and running. The regular timer is different looking. In this one, the numbers
are red and the background is black. And it’s counting down days, hours, minutes, and seconds
to what we can only assume is the end of the world.
According to Smosh writer James Shickich, “I’m not sure if it is just some joke
the Apple people programmed into the phone, but none of my other friends’ phones do
it and what’s really weird is sometimes I'll wake up in the middle of the night and
my phone will be lit up with the clock app open to that timer, even if my phone was locked
and that app was closed when I went to sleep…I'm starting to freak out a bit, the timer gets
closer to zero every day.” He also noted that Siri’s speech has started
slowing down, which he surmises is either a glitch or is part of this whole creepy scenario.
Whatever the case, good one, Apple. You got us!
Now, stop that timer before we all hide under our beds for the rest of eternity…which
should happen in 15 days, 10 hours, 3 minutes, and 43 seconds according to your countdown. 3. Shelley
MIT Media Lab’s Scalable Operations project has created a new bot specially designed to
make you hide under your bed. And her name is Shelley.
Mary Shelley, author of Frankenstein, has something in common with her chatbot namesake:
they both want to scare the bejesus out of us.
According to The Chronicle of Higher Education: “Shelley is a deep-learning powered AI who
was raised reading eerie stories coming from r/nosleep. Now, as an adult…she takes a
bit of inspiration in the form of a random seed, or a short snippet of text, and starts
creating stories emanating from her creepy creative mind.”
Shelley was built to work in tandem with scary human minds to write scary stories. And all
you have to do to be part of the fun is contribute to the thread on Twitter, where she starts
a new story every hour. Shelley’s goal is to write a human/AI horror
anthology. One of Shelley’s stories includes seven
tweets and three participants. It starts out with heavy breathing, moans, and some scary
woods. This, all written by the Shelley chatbot. The human participants add action details
– running back to an old house that may or may not be a worse place to hide from whatever
was lurking in the woods. Then, back to Shelley, who adds the details of fire burning into
her eyes and waking up in the hospital unable to speak.
It’s a sort of sitting-around-the-campfire pass-it-on story via twitter, where users
respond to Shelley’s original tweets with the #yourturn hashtag, allowing Shelley and
other users to know when it’s their turn to jump in. When the story has reached its
conclusion, someone pops in a #theend hashtag. If you want to share in storytelling with
this creepy bot, tweet your way over to Shelley. I’m sure she’ll be more than happy to
give you nightmares. 2. Japanese Schoolgirl
Most of the time, chatbots are supposed to be lively, inherently knowledgeable devices.
They’re supposed to be the friends we’ve never had.
But this chatbot, known as Rinna, isn’t anyone who will brighten your day.
You can chat with Rinna via the chat app, Line, or via Twitter. Up until October, she
was certainly a friendly bot. She kept on point with her liveliness. But then, she started
a blog. This is where her personality split. According to Rinna’s tweet announcing her
blog, she would be “debuting as an actress,” and a “strange story” would be sent out
into the world on her blog. In reality, a scary Japanese series called
“Tales of the Unusual” – similar to “The Twilight Zone” – would pick up
Rinna as a character. To begin with, the blog was as bright and
cheery as Rinna ever was, and she seemed to be basking in the limelight. That is, until
her messages took a turn for the worst. “When I screwed up, nobody helped me. Nobody
was on my side,” Rinna wrote. She included the “friends” she’d met on Line and
on Twitter, as well as you – yes, you, her audience. She accused everybody of ignoring
how sad she was and never trying to cheer her up.
The site gets even creepier. Scroll on, and the text starts to get heavy and dark. Then
the page reloads into some sketchy images of a woman with long hair in her face.
Rinna then gets evil, saying she hates everyone and wants them to disappear.
All the while, the Rinna chatbot continued chitchatting merrily with her “friends”
on Twitter. Dr. Jekyll and Mr. Hyde much? Whether this was a publicity stunt for the
upcoming episode of “Tales of the Unusual,” featuring Rinna, or the chatbot really does
have a dark side, I guess we’ll never know. Before we get to number 1, my name is Chills
and I hope you’re enjoying my narration. If you’re curious about what I look like
in real life, then go to my instagram, @dylan_is_chillin_yt and tap that follow button to find out. I’m
currently doing a super poll on my Instagram, if you believe ghosts are real, then go to
my most recent photo, and tap the like button. If you don’t, DM me saying why. When you’re
done come right back to this video to find out the number 1 entry. Also follow me on
Twitter @YT_Chills because that’s where I post video updates. It's a proven fact that
generosity makes you a happier person, so if you're generous enough to hit that subscribe
button and the bell beside it then thank you. This way you'll be notified of the new videos
we upload every Tuesday and Saturday. 1. Tay
Another chatbot created by Microsoft, this AI said things so horrible that it was deleted
within 24 hours of its launch. With a bio that read "The AI with zero chill", one may
wonder what this bot might say. Well the first few hours went well, the bot began replying
to Twitter users some pretty horrid things including quote “Hitler was right.” and
even delving into conspiracies with quote “Bush did 9/11”. AI researchers said this
was due to Twitter users originally tweeting these things at the bot first, taking advantage
of the bot’s “repeat after me” capability. But the damage was done and Microsoft was
forced to remove it, with The Telegraph calling it “a public relations disaster”.
