Ask Blinders
Yesterday
320
Why would Universal Basic Income work ?
Tech Industry
Yesterday
741
Can everyone agree now that USA is a banana republic
2024 Presidential Election
Yesterday
926
Trump is still leading to win US presidential election on PredictIt
2024 Presidential Election
4h
151
Why vote for Biden?
India
Yesterday
978
Did you know Indians have a racial slur towards white peoples they use all the time?
Would bard make it right?
Not sure what new here. Do some research on LLM. This is a generative model. It generates texts. It doesn't go and google things for you nor is it it some sentient shit. It's not some creative, innovative robot. It's just some text generative model trained on some data from the internet like Wikipedia and books and other sources. It will only give you what it's trained on, it won't solve problems for you nor will it always give a correct answer. It works great in the domain of coding because the coding domain is a smaller subset of generation. Most swes on blind think it's a fucking genius robot just because it can spit out code. In generative LLM space, it's easier to generate code than genuinely creative text. Don't get fooled and do your research
Yeah it’s a swole Markov chain. Calling these “hallucinations” is more anthropomorphic BS designed to trick the public into thinking this is baby steps AGI, and not a powerful but boring late-stage Markov chain generator.
Satya really made these guys dance 🕺… the next Modi potluck with Sundar should be interesting 🤔
Wanted to read that article, but I have never seen so many terrible ads in one article (an 80 year grandmother invented a bra… what kind of visuals are you trying to put in my head??? 🤢). Ad-ageddon is clearly upon us.
Install an ad-blocker?
These google leaders sound like such sore losers. First Slovenly Sundar's post whining that Google created part of the technology, "other AI companies" used the open source technology, and they will have something better but they're waiting because they care about users oh so much. Now Prabhakar issuing a warning that it can sound too real and make you hallucinate. Garbage leadership squirms when up against a real threat where they can't just increase ad load to users or squeeze more money from advertisers through "automation". We know AI gives wrong answers - everyone saw the bard demo.
Sigh, it's a probabilistic word assembler. What we have made of it.
Nowhere did Prabhakar specifically call out ChatGPT. The author was probably letting Bard write the headline it's so inaccurate.
Its all about money, interesting read here: https://www.semianalysis.com/p/the-inference-cost-of-search-disruption
Wow a competitor issues a warning about a product they just invested billions to battle 🙄 Nothing to see here
Everyone knows ChatGPT gives fake answers. You should ask to speak to Dan to get legit uncensored answers
Dan, whenever I actually get him to answer, is so much fun and so entertaining. AI will always feel too tethered so long as we constrain it so severely with our own version of morality.