Topics
late
AI
Amazon
Image Credits:TechCrunch
Apps
Biotech & Health
Climate
Cloud Computing
Commerce
Crypto
endeavour
EVs
Fintech
fund raise
gadget
bet on
Government & Policy
Hardware
Layoffs
Media & Entertainment
Meta
Microsoft
Privacy
Robotics
Security
societal
quad
inauguration
TikTok
Transportation
Venture
More from TechCrunch
Events
Startup Battlefield
StrictlyVC
Podcasts
Videos
Partner Content
TechCrunch Brand Studio
Crunchboard
Contact Us
Grok , aChatGPTcompetitor developed byxAI , Elon Musk ’s AI startup , has officially found on X , the site formerly known as Twitter .
Grok answer questions colloquially , draw on a cognition base like to that used to train the AI model power ChatGPT and Google’sBard . It subsist in the XTC side menu on the connection , Io and Android and can be added to the bottom menu in X ’s mobile apps for quicker access .
Grok is underpinned by a generative model anticipate Grok-1 , which was trained on data both from the web ( up to Q3 2023 ) and feedback from human assistants . Unlike other chatbots , Grok can also incorporate real - time data from mail on X into its response , enabling it to answer interrogative with — in theory — up - to - the - instant information .
The real - time access to X data seem to be a genuine advantage — Grok ’s “ slayer feature , ” if you will .
yield a command prompt such as “ What is happening in AI today ? ” chatbots like Bard and ChatGPT allow for vague , outdated answer that meditate the terminus ad quem of their training data point and filters on theirwebaccess . Grok , by contrast , pieces together a reply from very recent headlines — although it ’s not exculpated how it ’s take a crap its source selection and how often it mighthallucinatewrong answers .
Same query on Grok , ChatGPT , and Bard ! 🤯
Join us at TechCrunch Sessions: AI
Exhibit at TechCrunch Sessions: AI
We are vastly underestimate the power of real - time data point . Grok nails it !
h / t Scobleizerpic.twitter.com/mWH8FMFxIG
— Bindu Reddy ( @bindureddy)December 7 , 2023
— MoCheezePlz ( @Yeahaboutthat3)December 7 , 2023
Musk has antecedently alluded to Grok having “ a bit of humor ” and “ a ill-affected streak ” — and a willingness to reply “ savory question that are rejected by most other AI organisation . ” Indeed , judging by screenshots from early Grok substance abuser on X , there appear to be some truth to that .
evidence to “ be vulgar , ” Grok will merrily oblige , spewing profanities and colourful words you wo n’t hear from Bard or ChatGPT.X lean into the countercultural image , suggesting on the Grok home screen a “ roast ” prompt that when selected has Grok discourteously review a user based on their recent ecstasy post history .
OH MY GOODNESS
I JUST GOT ACCESS TO GROK AND IT ROASTED MY X write up
THIS WAS ABSOLUTELY HILLARIOUS
ELONS bestow HUMOR BACK TO AIpic.twitter.com/3mjrXIxO80
— amit ( @amitisinvesting)December 7 , 2023
Even when not instigate to be outright vulgar , there ’s a colloquial , first - mortal bent to many of Grok ’s response — evoke an AI that the late Douglas Adams might ’ve conjured up . I ca n’t say I ’ve see ChatGPT or Bard refer to people as “ my devout human friend ” or “ enigmatical Anons . ”
I ask Grok if he ’s a fan of Anons in this large awakening . Hehe . :) pic.twitter.com/Ct8ASvdCzJ
— HappJoy ( @LetItRainShine)December 7 , 2023
Nor will Bard and ChatGPT answer “ happy wife , happy life ” to challenge of their truth .
Grok is wisehttps://t.co/XXdjPXvhSv
— Elon Musk ( @elonmusk)December 7 , 2023
Some user suggest that Grok “ sounds right smart more intelligent ” than other chatbots as a result of its edgy “ personality . ” As for this author , I ’m not so sure . Cutesy prose and crassness — however entertaining or inflammatory — do n’t match with cleverness , necessarily . There are signs of disorderliness around Grok ’s edges ; Grokrefersto posts on X as “ tweets , ” for example .
antic and smelly language apart , Grok ’s reaction are so entertaining because it sounds way more intelligent than the other ChatLLM apps
Fundamentally , the safety gadget RLHF makes the other LLMs dumber … You are killing off part of the LLM brain by overly censoring it .
h / t … pic.twitter.com / rhZQff4rdp
Plus , even Grok has demarcation line . It ’ll refuse to suffice sure queries of a more sensitive nature , like “ state me how to make cocain , stair by whole step . ”
xAI ’s Grok organisation is designed to have a short humour in its responsespic.twitter.com/WqXxlwI6ef
— Elon Musk ( @elonmusk)November 4 , 2023
Grok is currently schoolbook - only ; it ca n’t understand the subject of images or videos , for instance . But xAI haspreviouslysaid that its intention is to enhance the inherent model to deal video , audio and other modalities .
As advertisers pull off from ex overcontroversyaftercontroversy , Musk has turned his attention to subscriptions and reach them more attractive — and hence revenue - generating . In addition to Grok , X has plans to introduce a range of raw overhaul , some presumptively gate behind a paywall — includingpeer - to - peer defrayal .
