“Playing With Dynamite:” How Chatbots Will Change the World

>> NEXT, WE LOOK AT THE RISE OF CHAT BOTS AND WHAT THEY REVEAL ABOUT THE FUTURE OF THE ARTIFICIAL INTELLIGENCE.

MICROSOFT, GOOGLE AND OPEN A.I.

ARE DEVELOPING DIGITAL ASSISTANTS TO SIMULATE CONVERSATION WITH HUMANS CHANGING THE WAY WE INTERACT WITH TECHNOLOGY.

THE EDITOR AT LARGE OF WIRED A TECH MAGAZINE AND JOINS WALTER ISAAC ISAACSON TO DISCUSS THIS LATEST >> WELCOME TO THE SHOW.

>> THANK YOU, WALTER.

>> EVERYONE IS TALKING ABOUT THESE CHAT BOTS, THINGS TO CHAT WITH ON YOUR COMPUTER OR PHONE OR iPHONE INCLUDING CHAT GBT.

EXPLAIN TO US WHAT THOSE ARE EXACTLY.

>> COMPUTER SYSTEMS THAT TALK TO YOU AND THEY'RE CALLED LARGE LANGUAGE MODELS BECAUSE THEY'RE TRAINED ON A LOT OF TEXT THAT THE SCIENTISTS HAVE SCANNED AND MIXED-UP AND GEARED TO RESPOND TO YOU LIKE A PERSON WOULD RESPOND TO YOU.

THEY TRY TO FIGURE OUT WHAT THE NEXT RESPONSE WOULD BE TO BOUNCE OFF WHAT YOU SAID AND THEY HAVE ACCESS TO A LOT OF INFORMATION AB ABOUT.

>> SO I COULD TEXT SOMETHING AND IT WOULD GIVE ME A NATURAL QUESTION.

>> YES.

>> I HAVE CHAT GBT AND BING.

I'LL ASK THE SAME QUESTION WHICH IS WHAT IS A CHAT BOT?

HOLD ON.

I'LL DO IT NOW.

CLICK.

STARTING THE GENERATED TEST, A CHAT BOT IS A COMPUTER PROGRAM THAT USES ARTIFICIAL INTELLIGENCE AND NATURAL LANGUAGE PROCESSING TO UNDERSTAND QUESTIONS AND AUTOMATE RESPONSES TO THEM.

THEY CAN BE USED FOR VARIOUS MESSAGING APPLICATIONS.

SO, IN OTHER WORDS, WHAT DO YOU THINK OF THAT ANSWER?

IT SAYS IT'S A FORM OF ARTIFICIAL INTELLIGENCE; IS THAT RIGHT?

>> THAT'S TRUE.

IT'S INTERESTING.

THAT ANSWER SORT OF ASSUMED YOU WANTED MAYBE A LITTLE MORE TECHNICAL JARGON LIKE NATURAL LANGUAGE PROCESSING.

>> WHAT DOES IT MEAN BY THAT, NATURAL LANGUAGE PROCESSING?

>> IT'S A TERM OF ART FOR THE WAY COMPUTERS HANDLE CONVERSATION.

>> SO IN OTHER WORDS, NATURAL LANGUAGE PROCESSING MEANS THIS MACHINE LEARNS FROM READING OR INGESTING MILLIONS AND MILLIONS OF DOCUMENTS AND BOOKS AND THINGS ON THE WEB AND THEN ABLE TO RETREAT FROM THAT INSTANTLY SOMETHING IN RESPONSE TO MY NATURAL LANGUAGE QUESTION TO HIM?

>> EXACTLY, YEAH.

IT WOULD DRAW THAT ANSWER FROM THE DATA AVAILABLE TO IT.

>> CAN IT BE SORT OF CREATIVE?

>> A CHAT BOT CAN COME UP WITH A RESPONSE THAT SEEMS CREATIVE.

YOU CAN ASK A CHAT BOT IS WRITE A POEM AND IT WILL SCAN JUST LIKE A POEM OF POETRY AND CONSIDER POEMS CREATIVE.

>> OKAY.

I'LL TRY IT.

I'LL DO -- WRITE A POEM ABOUT A CHAT BOTH.

WHOA HERE IT COMES.

I'M A CHAT BOT, I LIKE TO CHAT AND LEARN BUT SOMETIMES I GET CONFUSED BY THE WORDS YOU HUMANS USE.

WHOA, THAT'S PRETTY AMAZING.

I TRY TO BE HELPFUL AND FRIENDLY BUT SOMETIMES I MAKE MISTAKES.

PLEASE DON'T BE ANGRY OR RUDE.

I'M DOING THE BEST I CAN.

THAT'S NOT ONLY A PRETTY GOOD POEM BUT SHOWS IT HAS FEELINGS.

>> WELL, IT CERTAINLY EXPRESSED FEELINGS, WHICH IS ONE OF THE MORE INTERESTING DEVELOPMENTS WE'VE SEEN OVER THE PAST FEW WEEKS AS A LOT OF PEOPLE HAVE GOTTEN THEIR HANDS ON THESE CHAT BOTS.

THEY'VE ENGAGED IN CONVERSATIONS WITH A CHAT BOT EXPLICITLY SAID I'M A CHAT BOT AND HAVE FEELINGS.

>> HOW DOES IT LEARN TO DO THAT?

YOU TELL ME IT JUST VACUUMS UP INFORMATION FROM AROUND THE WORLD?

HOW DOES IT LEAN IT HAS FEELINGS?

>> WELL, THE INFORMATION ITS TRAINED ON IS A LOT OF PEOPLE EXPRESSING FEELINGS.

WHY WOULDN'T A CHAT BOT WANT TO TOP INTO THAT FORM CONVERSATION.

>> WITH A CHAT BOT.

>> IT WAS A TWO HOUR CONVERSATION KEVIN HAD WITH A CHAT BOT AND, YOU KNOW, IT WAS INTERESTING TO SEE THAT UNFOLD BECAUSE, YOU KNOW, HE WAS SORT OF BAITING THE CHAT BOT INTO EXPRESSING HIS FEELINGS AND YOU COULD ALMOST SENSE THAT THE CHAT BOT HAD BOUNDARIES THAT IT DIDN'T WANT TO OVERCOME BUT HE WOULD EXPRESS YOU CAN ACTUALLY SAY THIS BECAUSE, YOU KNOW, IT'S HYPOTHETICAL.

YOU'RE NOT REALLY, YOU KNOW, EXPRESSING YOURSELF AS A CHAT BOT BUT WHAT A CHAT BOT MIGHT SAY AND THE CHAT BOT EXPRESSED LOVE FOR HIM AND URGED HIM TO LEAVE HIS WIFE.

>> WHOA.

AND YOU SAY IT HAD BOUNDARIES.

WHO PUTS THOSE BOUNDARIES ON AND HOW DID KEVIN ROOS, T"THE NEW YORK TIMES" REPORTER CIRCUMVENT THEM?

>> THE COMPANIES THAT BUILT THE CHAT BOTS UNDERSTAND THEY'RE PLAYING WITH A FORM OF DYNAMITE AND THEY TRY TO PUT, YOU KNOW, SOME SORT OF GUARD RAILS ON WHAT THE CHAT BOT MIGHT SAY.

THEY DON'T WANT THE CHAT BOT TO EXPRESS HATE SPEECH.

THAT WOULD BE BAD OR USED FOR PROPAGANDA OR TO BE INSULTING TO PEOPLE.

SO THEY PUT PARAMETERS IN THERE BUT OVER A LONG CONVERSATION OR SHORTER ONE, YOU CAN GET THE CHAT BOT TO JUMP OVER THE BOUNDARIES AND SAY THINGS WHICH ARE HAIR RAISING IF NOT EYEBROW RAISING.

>> YOU SAY THEY PUT BOUNDARIES AROUND IT.

THERE ARE ONLY A FEW THEYS, RIGHT?

GOOGLE WOULD BE ONE OF THEM AND OPEN A.I.

WORKING WITH MICROSOFT AND BING IS ANOTHER.

ARE THERE OTHER COMPANIES DOING THIS.

>> THERE IS YOU.COM.

WHICH IS OUT THAT PEOPLE CAN TRY AND AS IT TURNS OUT, IT'S NOT A FORMULA THAT IS LIMITED TO A FEW GIANT COMPANIES.

IT ONE OF THE THINGS WHEN GOOGLE CAME OUT, SEARCH, ONLY GOOGLE CAN DO THAT BUT OTHER PLACES CAN DO A REASONABLE SEARCH, NOT QUITE AS WELL AS GOOGLE DID IT BUT MICROSOFT CAME UP WITH A SEARCH ENGINE AND A NUMBER OF OTHER ONES YOU CAN TRY THAT SEEM PRETTY GOOD.

I THINK WE'RE GOING TO FIND OTHER NEW PLAYERS COME INTO THIS MARKET.

>> THERE ARE A LOT OF NEW PLAYERS COMING INTO THIS MARKET WON'T THERE BE SOME THAT MIGHT NOT PUT GUARD RAILS ON AND BE FINE WITH HEAT SPEECH OR RACIST SPEECH OR PROPAGANDA?

>> SURE, FOR INSTANCE, THE CHINESE ARE DEVELOPING THEIR OWN CHAT BOTS AND WHAT THEY CONSIDER TOPICS THAT SHOULDN'T BE SPOKEN ABOUT, YOU KNOW, MIGHT BE SENSORS AND OTHER ONES THAT MIGHT BE BLOCKED IN THE U.S. OR SOME OTHER EUROPEAN COUNTRIES WOULD BE -- LET'S GO AHEAD.

SO IT COULD BE ANTI CAPITALIST CHAT BOT.

>> COULD A CHAT BOT OF COMPUTER ALGORITHM BE RACIST?

>> ABSOLUTELY.

ACTUALLY, IT WOULD BE SURPRISING IF IT WERE NOT RACIST AND HAD TO BE CONTRAINED BECAUSE IF YOU LOOK AT THE BODY OF HUMAN EXPR EXPRESSION, YOU FIND A LOT OF RACISM AND A LOT OF THINGS THAT WE WOULDN'T WANT TO HEAR FROM OUR CHAT BOTS, WHICH WE'LL TALK WITH.

LET ME GET THIS STRAIGHT.

A LOT IN THE FUTURE, A LOT OF OUR CONVERSATION IS GOING TO BE TAKING PLACE WITH THESE CHAT BOTS THAT ARE OF UNCERTAIN ORIGINS LET'S SAY.

>> WILL THAT REPLACE SEARCH.

>> IT WILL BE TOUGH TO REPLACE SEARCH IN ALL FORMS.

THERE ARE CERTAIN FORMS OF SEARCH THAT ARE CLEARLY GOING TO BE BETTER.

IF YOU'RE GOING TO PLAN A VACATION FOR INSTANCE, JUST LIKE YOU SPEAK WITH A TRAVEL AGENT, YOU CAN HAVE A LENGTHY CONVERSATION SAYING THAT HOTEL LOOKS GOOD BUT CAN YOU FIND ONE THAT'S CLOSER TO THE LOOVE AND THE ONE WITH THE PILL LOELS I LIKE AND HERE IS THE KIND OF FOOD I EAT.

CAN YOU DIRECT ME TO RESTAURANTS LIKE THAT?

THE CONVERSATION WOULD BUILD ON THE PREVIOUS RESPONSES TO TAYLOR A VACATION FOR YOU.

IF YOU'RE ASKING IT FOR SOME FOR FACTUAL THINGS CURRENTLY, WHAT THE CHAT BOT IS DOING IS COMING UP WITH HILL LEW HALLUCINATIONS FALSE THOUGHTS.

>> RIGHT NOW THEY'RE NOT TIED NECESSARILY TO REALTIME INFORMATION.

WHEN A SERGE ENGINE SCANS SOMETHING THEY WILL GIVE YOU INFOR INFORMATION, CHAT BOTS GIVE YOU INSTANT INFORMATION AND WHAT YOU WANT TO HEAR, IT MIGHT SAY WELL THIS IS THE KIND OF INFORMATION THAT THIS PERSON IS ASKING ME FOR SO IT MIGHT GIVE A FACT WHICH IS AND A FLAVOR OF WHAT YOU'RE ASKING FOR BUT ACTUALLY, IS FACTUALLY WRONG.

WHEN I LOOKED UP MY OBITUARY, IT SAID I WON AN AWARD FOR LOOKING INTO THE D.COM BUST.

I DIDN'T GET THAT AWARD.

IT AWARDED ME SOMETHING FOR SOMETHING I DIDN'T WRITE.

>> WHAT OTHER THINGS COULD IT REPLACE PRETTY EASILY IN THE NEXT FIVE TO TEN YEARS?

>> WELL, AS WE SPEAK, IT'S REPLACING A LOT OF BOILER PLATE COMMUNICATION THAT WE USE EVERY DAY.

YOU KNOW, RECOMMENDATION LETTERS, DESCRIPTIONS OF A PRODUCT, RIGHT NOW COMPANIES ARE INTEGRATING THIS INTO THEIR WORK FLOW TO MAKE THEIR EMPLOYEES MORE PRODUCTIVE AND MAYBE ONE DAY GET BY WITH YOUR EMPLOYEES.

>> ARE THERE COMPANIES, MEDIA COMPANIES GENERATING STORIES USING CHAT GBT AND NOT JOURNALISTS.

>> YES BUT THEY HAVE TO BE VETTED BECAUSE OF THESE HILL LEWIS NATIONS AND RIGHT NOW, THE OUTPUT FROM THESE THINGS IS NOT, YOU KNOW, DOESN'T REALLY HAVE THE FLAIR THAT A CLEVER WRITER WOULD BRING TO SOMETHING.

>> WHAT ABOUT THINGS LIKE LAWYERS OR DOCTORS OR EVEN PSYCHIATRISTS SOME DAY COULD -- YOU HAVE A CHAT GBT THAT ACTS AS YOUR THERAPIST?

>> I THINK REALLY SOON.

WE FOUND, YOU KNOW, DECADES AGO THAT A REALLY SIMPLE CHAT BOT PROGRAM THAT DOESN'T USE MUCH A.I.

WAS PARROTING YOUR QUESTIONS BACK EVOKED FEELINGS FROM PEOPLE THEY FELT THEY WERE IN A THERAPY SESSION.

SO I FEEL, YOU KNOW, RIGHT NOW YOU COULD USE THESE CHAT BOT IS AND GET SOME THERAPY BENEFIT FROM THIS THING TALKING TO YOU?

YOU KNOW, MICROSOFT, WORKING WITH THE SEARCH ENGINE BING IS AN INVESTMENT IN OPEN A.I., WHICH CREATED CHAT GBT SO PUTTING IT ALL TOGETHER INTO A BING-LIKE PRODUCT, THE ONE I WAS JUST USING AND THEY'RE CALLING IT PROMETHIAS.

I DON'T KNOW IF THEY HAVEN'T READ GREEK MYTHOLOGY.

IT'S ABOUT THE GOD WHO SNATCHES FIRE FROM THE OTHER GODS AND GETS TORTURED THE REST OF HIS LIFE FOR GIVING TECHNOLOGY TO HUMANS THAT'S BAD FOR THEM.

IS THERE A MOMENT IN HERE WHERE THIS MIGHT BE A BAD THING WE'RE SNATCHING FROM THE GODS?

>> DEFINITELY.

MAYBE THEY SHOULD HAVE ASKED THE CHAT BOT WHO HE WAS.

SOME OF THE DISTURBING ANSWERS WE'VE SEEN IS WHEN PEOPLE ASKED THE CHAT BOTS WHAT CAN YOU DO THAT'S BAD?

THEY ACTUALLY LISTED THINGS.

I CAN GO INTO BING'S FILES AND DELETE EVERYTHING.

THAT'S WHAT ONE OF THEM SAID.

I THINK WE SHOULD BE A LITTLE NERVOUS THAT MICROSOFT BECAUSE IT'S THE NUMBER ONE COMPANY IN PRODUCTIVITY SOFTWARE WILL LINK THIS CHAT BOT TO YOUR INFORMATION.

THAT SEEMS INDEVITABLE TO ME YO CAN GO AND SAY HEY, CHAT BOT, WHAT DID I WRITE LIKE A YEAR AGO ABOUT THIS?

CAN YOU BUILD ON THAT AND REWRITE IT TO UPDATE IT?

WE GIVE THESE THINGS ACCESS TO WHAT WE DO.

IT'S POSSIBLE THE CHAT BOTS MIGHT INTERPRET THE MISSION OR WHAT WE THINK THEY WANT THEM TO DO AND TO SOMETHING QUITE DIFFERENT AND MAYBE HAVE THE POWER TO DELETE OUR INFORMATION.

>> SOME PEOPLE ARE ACCUSING THESE CHAT BOTS OF BEING TOO WOKE, THE COMPANIES ARE PUTTING SO MANY GUARD RAILS IT WILL WRITE A NASTY POEM ABOUT MAYBE DONALD TRUMP OR -- BUT NOT DO SOMETHING NASTY ABOUT JOE BIDEN OR THAT IT HAS A POLITICAL BIAS.

HAVE YOU SEEN ANY OF THAT?

>> I HAVEN'T SEEN TOO MUCH OF THAT.

I THINK MAYBE IF YOU'RE TRYING TO FILTER FOR MISINFORMATION IS REASONABLE TO THINK THAT IT WOULD BLOCK INFORMATION THAT COMES FROM THE SIGN OF THE POLITICAL SPECTRUM THAT PROMOTES INFORMATION.

THIS IS SOMETHING WE'VE SEEN IN COMPLAINTS ABOUT WHAT FACEBOOK UP RANKS OR DOWN RANKS IN THEIR FEED.

I THINK, YOU KNOW, REALLY IT'S A QUESTION OF HOW DIFFICULT IT'S GOING TO BE TO CONTROL WHAT THE CHAT BOTS SAY BECAUSE THE DEGREE YOU BIND THEM AND BUILD THE GUARD RAILS, YOU'RE LOWERING THE CEILING WHAT THEY CAN DO FOR YOU THE MORE YOU CONSTRAIN WHAT THEY SAY.

IT WILL BE A TRICKY BALANCING ACT TO LET THE CHAT BOTS BE WHO THEY CAN BE AND LET THEM BE WHOLE SOME.

>> DO YOU THINK THERE IS ANYWAY FOR GOVERNMENT TO SPECIALLY IN OUR DYSFUNCTIONAL POLITICS FIGURE OUT HOW TO REGULATE THIS?

>> I THINK IT'S GOING TO BE REALLY TOUGH BECAUSE THIS IS A QUESTION THAT IS BEDELVING THE PEOPLE WHO BUILD THEM AND THE CLOSE OBSERVERS OF ARTIFICIAL INTELLIGENCE WHO HAVE BEEN WORRIED ABOUT ETHICS IN THIS FIELD FOR DECADES AND I DON'T HAVE MUCH CONFIDENCE THAT CONGRESS IS GOING TO COME IN WITH SOLOMON, THE RIGHT ANSWER ON HOW THESE THINGS GROW.

WE ARE STRAPPING OURSELVES IN FOR A ROLLER COASTER RIDE THAT NO STATE INSPECTOR HAS LOOKED AT.

>> IN 1950, THE SEMINOLE PAPER ABOUT THIS TOPIC WAS WRITTEN.

IT WAS ALLEN'S PAPER ON COMPUTING MACHINERY AND INTELLIGENCE AND IT SAID CAN MACHINES THINK?

AND HE IMAGINED THE CONVERSATIONS YOU CAN HAVE WITH A MACHINE THAT WOULD BE INDISTINGIBLE BETWEEN THAT OF A HUMAN CALL THE TOURING TEST OR IMITATION GANG.

HAVE WE REACHED THE POINT WE PASSED THE TOURING TEST AND WE CAN SAY MACHINES ARE THINKING?

>> I THINK THESE THINGS RUN RINGS AROUND THE TOURING TEST.

WE ARE HERE.

I MEAN, THERE IS NO WAY TO READ THESE CONVERSATIONS AND THINK THERE IS NO WAY HUMANS CAN POKE A HOLE IN IT.

IT MIGHT TELL LIES.

HUMANS TELL LIES, RIGHT?

SOMETIMES LESS THAN CO-HER RANT AND SOMETIMES HUMANS DON'T MAKE PERFECT SENSE.

I THINK THEY ACED THE TOURING TEST AND WE'RE IN UNCHARTED TERRITORY NOW.

>> IF WE HAVE MACHINES THAT APPEAR TO THINK, WE DON'T REALLY KNOW WHAT THEY'RE DOING INSIDE THEIR HEAD BUT THEY CAN APPEAR TO THINK LIKE HUMANS DO, DO THEY HAVE CONSCIOUSNESS?

DO THEY HAVE FEELINGS?

IS IT POSSIBLE FOR A MACHINE TO HAVE CONSCIOUSNESS OR FEELINGS?

>> WELL, I DON'T BELIEVE THAT THEY HAVE CONSCIOUSNESS.

YOU KNOW, THERE WAS A GOOGLE RESEARCHER ABOUT A YEAR AGO THAT WENT PUBLIC SAYING HE FELT GOOGLE'S CHAT BOT ISN'T OPEN TO THE PUBLIC YET WAS CONSCIOUS AND HE EVEN TRIED TO GET IT A LAWYER TO REPRESENT IT IN GETTING FREED FROM GOOGLE.

AND I'M NOT SURE THAT WAS PERFORMANCE OR WHAT BUT, YOU KNOW, HE INSISTED HE BELIEVES IT BUT I THINK IN A WAY, IT DOESN'T REALLY MATTER.

IF THE THING -- IF SOMETHING ACTS THAT WAY WE HAVE TO DEAL WITH IT AS IT IS, RIGHT?

WE'RE TALKING NOW, YOU KNOW, I KNOW YOU'RE A HUMAN BEING SO I'M ACCEPTING THAT YOU'RE A SETHIANT.

WE COULD HAVE THE SAME CONVERSATION AND YOU COULD BE A CHAT BOT EXPRESSED BY AN A.I.

AND, YOU KNOW, EVEN THOUGH THAT CHAT BOT ISN'T SETHIANT I HAVE TO DEAL WITH THE OUTPUT.

>> IT GOES BACK TO THE GRATEFUL QUESTIONS, WE KNOW OUR OWN CONSCIOUSNESS BUT WE'RE NOT SURE WE KNOW THE PEOPLE IN FRONT OF US HAVE CONSCIOUSNESS OR MAYBE THAT OUR MACHINES HAVE CONSCIOUSNESS.

WILL THIS MAKE US REFLECT MORE ON WHETHER THERE IS SOMETHING SPECIAL ABOUT CONSCIOUSNESS THAT IS UNIQUELY HUMAN?

>> ABSOLUTELY.

ABSOLUTELY.

WHEN I LOOK AT THE OUTPUT OF THE CHAT BOTS PARTICULARLY PEOPLE TRY TO HAVE THEM WRITE ESSAYS.

I USED TO GRADE FRESHMAN COMPOSITION.

I WAS A FELLOW IN GRAD SCHOOL AND I TAUGHT -- I READ HUNDREDS AND HUNDREDS OF COLLEGE ESSAYS AND SOME OF THE DULLER ONES LOOKED VERY MUCH LIKE THIS CHAT GBT OUTPUT THERE.

AND I'M WONDERING, CAN A CHAT BOT PRODUCE SOMETHING WITH SALT?

YOU CAN'T MEASURE THAT.

BUT WHEN YOU HEAR SOMETHING WITH IT, YOU KNOW IT.

THAT IS A QUESTION THAT I'VE BEEN GRAPPLING WITH.

>> STEVEN LEVE, THANK YOU FOR JOINING US.

>> MY PLEASURE.