i think it´s great that @grok 4.7 was being trained Specifically for Grok Bot (or at least that´s what it is really good at)
IF the Grok Bot usage gets better and they maybe hand out a reset for Grok Bot as an excuse for the model delay, i´d consider this a good model. Not great, but good
Grok 4.7 just released and it's an EXCELLENT model
It was trained FOR Grok Bot
Meaning this is a fully agentic model trained to do your knowledge work better than you can
In this video I show you how to use Grok 4.7 and a Grok Bot workflow that will 10x your productivity:
Grok 4.7 just released and it's an EXCELLENT model
It was trained FOR Grok Bot
Meaning this is a fully agentic model trained to do your knowledge work better than you can
In this video I show you how to use Grok 4.7 and a Grok Bot workflow that will 10x your productivity:
People saying Astra (or even Sol) is bad at Designing honestly just sucks at prompting.
Yes, you have to be more specific with what you want, but the results for me are better than fable tbh.
And no „make it real good“ is not a good prompt
It's a classic Elon move to promise something Or come up with a super ambitious timeline/release date. If it works, everyone is super, super happy and loves him because he's able to move just so fast.
But if it doesn't work, obviously people are a bit upset because things have to be rescheduled, Which is fine, but I would just love to have a little orientation.
Then I feel like he played Grok 4.7 a bit down over a period of time? Don't get me wrong, I'm super excited, but these things just mess up XAI's quality a little bit, if you get what I mean.
But yeah, overall I get what you are saying, 100%
@bot is probably the coolest thing i have ever used. It makes it so easy to organize everything and to build fast. I hope that is just gets better, once Grok 4.7 finally comes out
@bot is probably the coolest thing i have ever used. It makes it so easy to organize everything and to build fast. I hope that is just gets better, once Grok 4.7 finally comes out
been in the indurtry for 8 years now, ofc i use AI a lot in my Job nowadays.
Been always into tech and think this is one of the biggest revoltions in tech ever (obv even effecting everybody´s life)
But when your first intuition is to insult people you don´t even know because of a x post that brings up a problem everybody´s been having and OpenAI´s even admitting, we all see who the "retard" is.
If you feel personally offended because of an x post about AI usage, pls for the love of god get a life.
@andrebckr@byteboxmediauk@thsottiaux another retard who is a "designer" or works in tech but whos entire personality would fall apart and wouldnt be able to do what he does now if ai goes down tomorrow, you guys need to learn your craft and use ai as a tool, not have it do everything for you. what if the bubble pops
@andrearslanian Astra doesn‘t work like fable. You can not give a vague prompt and it‘ll design something good. You have to way more specific in what you want. But if you do that, the results feel better than fable in my opinion. You people just need to be specific, prompting is still a thing.
Like i said 4h ago, doing my job for 8 years, before AI was a thing. Can do without just as fine, takes longer of course.
Insulting people on a x post on AI and usage, losing the argument and not being able to admit the mistale should be the wake up call big dawg
You should learn basic human behaviour, communication, and putting things into context before commenting someone‘s post again
@dragocz95@thsottiaux same. I even switched completely to OpenAI from Anthropic because their usage sucked + the communication was off - i am sure OpenAi will handle this better but for now it kinda sucks
i mean 2 days out of 30. Of course there are resets, but usage got so unbelievably bad expecially with Astra, that most people can´t even use a mix of Astra and Sol/Terra/Luna.
Lots and lots of people i talk to run into the same problem, even when trying to reduce token usage as much as possible.
We don´t see resets as frequent with astra as they were with sol and i get that, really. It is an hyper intelligent model that need ton of compute. But guess why they provided so many resets? Because usage didn´t match your billing plan.
Tibo admitted that, and there is nothing wrong with that, i know they´ll fix it, but for now it kind of sucks seeing usage getting worse by the day
@bridgemindai i believe that this is the lesson of the day. "smarter" is unfortunately not always better. It makes so much fun using these models, but compute has to catch up first.
Probably only with Grok 5 will be able to have a hyperintelligent model + is able to provide the usage for it
@izypublic Yeah and i know a lot of people who are also doing token reduction methods and still only have a day??
„I know people“ has to be the best argument ever