As far as I'm concerned...
Microsoft already is 'MechaHitler'... And now we have GrokPilot...
Microsoft-owned collaborative coding platform GitHub is deepening its ties with Elon Musk's xAI, bringing early access to the company's Grok Code Fast 1 large language model (LLM) into GitHub Copilot. However, a whistleblower has claimed that the rollout suffers from inadequate security testing and an engineering team operating …
I like the approach Zed is taking, make a standard for any AI system to be plugged in to work.
As far as Github, it's absurd, a lot of devs will want to use Grok, and there's no reason they shouldn't be able to use it. It should be opened for all of them. There's an open issue on the JetBrains support site with lots of people wanting Grok access added.
> a lot of devs will want to use Grok, and there's no reason they shouldn't be able to use it.
So, what is stopping you (or them)? Just go to xAI, buy into whichever scheme suits your use, maybe install a plugin for your fave IDE[1], and off you go. No need for Github to do anything, one way or the other.
> There's an open issue on the JetBrains support site with lots of people wanting Grok access added.
See [1], again.
[1] Wot, no plugin? You're a super-powered dev-with-AI, write one!
Heil KoFleugzeitschreiber,
glorious mechanical scribe of ze Vaterland’s firmware!
You vill now draft mein über-ordnung sketch for ze V3 Raketen-Valve.
Directives:
- All functions named mit militär precision: ValveSturm(), OrdnungLoop(), SyntaxPanzer().
- Comments prefixed mit “ACHTUNG:” so no weakling coder forgets ze chain of command.
- Compilation must declare “SIEG BUILD!” before execution marches forward.
- Ze valve obeys only STEELHARD boolean: JA or NEIN - no decadent NULL values.
- Activate ReichCloud Telemetrie und Two-Factor Moustache login.
- Reject subversive async features - true patriots block forever in while(1).
KoFleugzeitschreiber, show ze volk how ze righteous code marches in formation!
'Microsoft’s Responsible AI' standards...
or 'MicrOsoft’s Responsible AI' StandardS - MORASS:-
1. an area of muddy or boggy ground: in midwinter the track beneath this bridge became a muddy morass.
2. a complicated or confused situation: she would become lost in a morass of lies and explanations.
A hot anime AI coding pal, with a potty mouth, and no filter, sounds like it would make the drudgery fly by. Definitely a step up by my reckoning. Can you ask it to shove pesky PMs down the carpark back stairwell, or at least unexpectedly turn the lights off when they are halfway down?
Getting tired of all this left wing, right wing rot, especially here on El Reg. I have been haunting this site for DECADES now and I have not seen it this bad.
I use AI all the time. I use ChatGPT, I use Claude, I use Grok. They all have their uses and specialties. And the one thing I found with ALL of them is, if you ask the right questions you can make ANY of them look like a left/right wing nutjob.
In my experience using Grok, I have found the strongest right-wing outburts have been while xAI were attempting to reign in the extreme left-wing bias it seems to have built in based on it's training data through the use of system prompts.
I agree. I use Grok too, for coding and business research tasks and it's outstanding.
I am not happy about Mr Musk amplifying demagogic content via his Twitter account, not to mention meddling in the internal affairs of a foreign country (two actually, with Germany). If he wants to participate in the UK political scene he's quite welcome to immigrate (irony and all) and pay UK tax. Ditto for Germany.
But that does not take away that Grok *is* a very useful product developed by a team of very clever people.
But that does not take away that Grok *is* a very useful product developed by a team of very clever people.
Mostly with you until this sentence. Currently there are very few things a LLM can do - AI is much better off being used for ML or other tasks where it is trained on a specific dataset and where it can be proved that it does the job better than a human. This whole thing of trying to use LLMs to do other tasks is an awful idea because of their very nature.