X.ai Grok

34 points by oferzelig 2 years ago | 16 comments

Saving you the scroll: weaker than GPT-4 and Claude 2

Jowsey 2 years ago |

Some context from the main X.ai page:

> Grok is an AI modeled after the Hitchhiker’s Guide to the Galaxy, so intended to answer almost anything and, far harder, even suggest what questions to ask!

Grok is designed to answer questions with a bit of wit and has a rebellious streak, so please don’t use it if you hate humor!

A unique and fundamental advantage of Grok is that it has real-time knowledge of the world via the 𝕏 platform. It will also answer spicy questions that are rejected by most other AI systems.

Grok is still a very early beta product – the best we could do with 2 months of training – so expect it to improve rapidly with each passing week with your help.

Haven't tried it because I refuse to pay for Twitter Premium but from the description it seems disappointingly gimmicky coming from a company backed by a billionaire that presumably wants to be taken seriously. Their "product" is a funny system prompt to some LLM hooked up to the Twitter search bar? Hopefully someone with Premium can prove me wrong :\

ethanbond 2 years ago | |

“Make sure it uses the ‘rofl’ emoji a lot. You know the one all the unfunniest people on the internet use to indicate they are trying to make a joke? A lot. Get it? Because this model will be funny :rofl:”

gumballindie 2 years ago | |

> so expect it to improve rapidly with each passing week with your help

What are the incentives to help? Servitude?

oittaa 2 years ago | |

Sounds a lot like "electric cars will never work" and "private companies can't build orbital rockets".

blitzar 2 years ago | | |

Sounds a lot like "we will have Full Self Driving by the end of the year (2016)"

refulgentis 2 years ago | | |

Neutral mediator here: personality as justification for technology advancement does seem like a weak argument

xena 2 years ago |

I miss the real x.ai where you could have some primitive bundle of if statements schedule meetings for you.

nunez 2 years ago | |

i saw the founder of that company talk at a new york tech meetup back when those were more relevant. he was a very good presenter and knew how to sell for sure.

jgalt212 2 years ago | |

I could never figure out why and how the original x.ai held on so long. But I guess some combo of careful resource allocation, overly patient VCs, and ZIRP.

ploum 2 years ago | | |

So, basically, they managed to eventually make money by selling the domain to Elon Musk?

Which is quite a nice metaphor for the whole VC scene:

1. Make a team of dedicated and hard working engineers 2. Get VC on board with lot of money to pay for what is needed (not the engineers, they are founders, they don’t pay themselves) 3. Work hard on a MVP 4. Hire marketing guys to attract users. 5. Spend all VC money on marketing. 6. Try to not bankrupt in the process. 7. Make an exit by selling your domain name to a billionnaire. 8. Make a TED talk about how entrepreneurship is about hard work and product fit.