Topics
Latest
AI
Amazon
Image Credits:OpenAI
Apps
Biotech & Health
Climate
Image Credits:OpenAI
Cloud Computing
Commerce Department
Crypto
Screenshot: ChatGPT transcribes the verbal conversation after.
go-ahead
EVs
Fintech
Fundraising
gismo
punt
Government & Policy
Hardware
layoff
Media & Entertainment
Meta
Microsoft
secrecy
Robotics
Security
Social
Space
Startups
TikTok
conveyance
speculation
More from TechCrunch
Events
Startup Battlefield
StrictlyVC
Podcasts
television
Partner Content
TechCrunch Brand Studio
Crunchboard
Contact Us
I ’ve been playing around withOpenAI ’s Advanced Voice Modefor the last workweek , and it ’s the most convincing taste I ’ve had of an AI - powered future yet . This week , my phone laughed at gag , made them back to me , asked me how my day was , and state me it ’s bear “ a great time . ” I was talking with my iPhone , not using it with my hands .
OpenAI ’s new feature article , presently in a circumscribed alpha trial , does n’t make ChatGPT any smarter than it was before . or else , Advanced Voice Mode ( AVM ) makes it friendlier and more innate to talk with . It make a novel user interface for using AI and your devices that feels fresh and exciting , and that ’s exactly what scare away me about it . The product was kinda glitchy , and the whole idea wholly creeps me out , but I was surprised by how much I genuinely enjoyed using it .
Taking a whole tone back , I think AVM fit into OpenAI CEO Sam Altman ’s tolerant visual modality , alongside agents , of changing the mode humans interact with figurer , with AI models front and center .
“ Eventually , you ’ll just ask the computer for what you need and it ’ll do all of these tasks for you , ” Altman say during OpenAI ’s Dev Day in November 2023 . “ These capabilities are often talked about in the AI field as ‘ agents . ’ The upside of this is going to be marvelous . ”
My friend, ChatGPT
On Wednesday , I tested the most grand upside for this advanced technology I could think of : I asked ChatGPT to order Taco Bell the fashion Obama would .
“ Uh , let me be absolved — I ’d like a Crunchwrap Supreme , maybe a few tacos for good measure , ” said ChatGPT ’s Advanced Voice Mode . “ How do you think he ’d treat the drive - thru ? ” tell ChatGPT , then laughing at its own joke .
The impression genuinely made me express joy as well , matching Obama ’s iconic cadence and pauses . That said , it stayed within the tone of the ChatGPT phonation I choose , Juniper , so that it would n’t be really confused with Obama ’s voice . It sounded like a friend doing a bad impression , infer exactly what I was trying to evoke from it , and even that it was saying something peculiar . I found it surprisingly joyful to talk with this advanced assistant in my phone .
Join us at TechCrunch Sessions: AI
Exhibit at TechCrunch Sessions: AI
I also expect ChatGPT for advice on navigate a problem involving complex human relationships : enquire a important other to move in with me . After explaining the complexity of the relationship and the counseling of our careers , I welcome some very detailed advice on how to progress . These are questions you could never ask Siri or Google Search , but now you could with ChatGPT . The chatbot ’s interpreter even show a slightly serious , aristocratic tone when react to these prompts ; a severe contrast from the joking note of Obama ’s Taco Bell club .
ChatGPT ’s AVM is also great for helping you understand complex case . I ask it to break down detail on an earnings story — such as free hard currency menstruum — in a agency that a 10 - year - old would sympathize . It used a lemonade stand as an example , and explained several fiscal terms in way my younger cousin would totally get . you could even ask ChatGPT ’s AVM to talk more tardily to meet you at your current level of understanding .
Siri walked so AVM could run
Compared to Siri or Alexa , ChatGPT ’s AVM is the clear-cut succeeder thanks to faster response times , unequalled answers , and its power to respond complex questions the anterior generation of practical assistants never could . However , AVM falls short in other means . ChatGPT ’s voice feature ca n’t set timekeeper or reminder , surf the web in real fourth dimension , check the weather , or interact with any genus Apis on your phone . Right now , at least , it ’s not an effectual replacement for virtual assistants .
Compared toGemini Live , Google ’s competing feature , AVM feel slightly ahead . Gemini Live ca n’t do impressions , does n’t express any emotion , ca n’t speed up or slow down , and takes longer to respond . Gemini Live does have more voice ( ten compared to OpenAI ’s four ) and seems to be more up to date ( Gemini Live lie with about Google ’s antitrust ruling ) . Notably , neither AVM nor Gemini Live will blab out , likely an effort to avoid run - Immigration and Naturalization Service with copyright lawsuit from the record industry .
That say , ChatGPT ’s AVM glitches a lot ( as does Gemini Live , to be mediocre ) . Sometimes it will cut itself short mid - judgment of conviction , then depart over . It also gets this uncanny , grainy - sounding vocalisation here and there that ’s a little unpleasant . I ’m not trusted if this is a trouble with the modelling , internet connexion , or something else , but these proficient shortcoming are somewhat expected for an alpha tryout . The problems did little to take me out of the experience of literally talking with my telephone set , though .
These examples , in my nous , are the beauty of AVM . The characteristic does n’t make ChatGPT all - knowing , but it does admit the great unwashed to interact with GPT-4o , the underlying AI modeling , in a unambiguously human elbow room . ( I ’d understand if you forgot there ’s no person on the other end of your phone . ) It almost feels like ChatGPT is socially aware when talking with AVM , but of course , it is not . It ’s simply a pile of neatly packaged predictive algorithms .
Talking tech
Frankly , the feature worries me . This is n’t the first prison term a engineering company has offered fellowship on your earphone . My propagation , Gen Z , was the first to maturate up alongside social media , where society offered connection but instead played with our collective insecurity . talk with an AI equipment — like what AVM appears to offer — seems to be the development of societal medium ’s “ friend in your phone ” phenomenon , offer up tatty connection that scratch at our human instincts . But this time , it removes humans from the loop completely .
contrived human connection has become asurprisingly pop use casefor generative AI . hoi polloi today are using AI chatbots as friend , mentor , healer , and teachers . When OpenAI launched its GPT store , it wasquickly flooded with “ AI girlfriend , ” chatbots specialized to move as your significant other . Two investigator from MIT Media Labissued a warningthis month to prepare for “ habit-forming news , ” or AI companions with sorry patterns to get humans hooked . We could be opening a Pandora ’s box seat for new , tantalizing ways for equipment to keep our attending .
in the beginning this month , a Harvard dropout shook the technology world by bait anAI necklace called Friend . The wearable gimmick — if it works as promised — is always listening , and the chatbot will text with you about your lifetime . While the musical theme seems crazy , innovations like ChatGPT ’s AVM gives me cause to take those purpose face seriously .
And while OpenAI is leading the boot here , Google is n’t far behind . I ’m confident Amazon and Apple are racing to put this capableness in their Cartesian product as well , and before long enough , it could become table bet for the manufacture .
think asking your smart TV for a hyper - specific recommendation for a movie , and get just that . Or tell Alexa exactly what cold symptom you ’re feeling , and in turn have it order you tissue and cough medicine on Amazon , while advising you on home remedies . Maybe you could ask your computer to draft a weekend trip for your family , or else of manually Googling everything .
Now , obviously , these natural action require boundary and saltation forwards in the AI agentive role world . OpenAI ’s endeavor on that front , the GPT store , feels like an overhyped product that ’s no longer much of a focal point for the company . But AVM at least take care of the “ talking to computers ” part of the puzzle . These concepts are a tenacious way out , but after using AVM , they seem a muckle near than they did last week .
