News for hackers. Stuff that matters.
Re:Grok3 Launch by dtquad on Tuesday February 18, 2025 @08:44AM (#43089366)
The full-sized DeepSeek-R1 is on par with o1.

o1-pro is "o1 on steroids" and was the first selling point of the $200/month Pro subscription but they later also added "Deep Research" and Operator to the Pro subscription.

Re:Grok3 Launch by guax on Tuesday February 18, 2025 @09:12AM (#43089678)
Every year seems like we get worse at naming things in non confusing ways. I am waiting for the o1-pro-max now, pro max ultra and pro max ultra plus.
Re:Grok3 Launch by jkestner on Tuesday February 18, 2025 @09:31AM (#43089893)
Then you’ll know what the naming AI was trained on.
Re:Grok3 Launch by theptip on Tuesday February 18, 2025 @12:59PM (#43092940)
It’s hard because there are multiple dimensions being upgraded at different cadences. Architecture, parameter count, etc.
Re:Grok3 Launch by polski-g on Tuesday February 18, 2025 @01:00PM (#43092958)
I was recently laid off from OpenAI. My job was coming up with names for their models.
Re:Grok3 Launch by meekaaku on Tuesday February 18, 2025 @02:46PM (#43094132)
Is this because of USAID funding cut?
Re:Grok3 Launch by rvnx on Tuesday February 18, 2025 @01:24PM (#43093235)
You are welcome to join the Bard team
Re:Grok3 Launch by fragmede on Tuesday February 18, 2025 @03:00PM (#43094292)
It's not a "layoff" if you were fired for poor performance/picking bad names.
Re:Grok3 Launch by AlchemistCamp on Tuesday February 18, 2025 @03:19PM (#43094513)
Microsoft already mastered this decades ago with a dozen different license tiers for Windows with unintuitive feature sets.
Re:Grok3 Launch by tartuffe78 on Wednesday February 19, 2025 @08:44AM (#43102062)
Not to mention the Xbox iterations...
Re:Grok3 Launch by guax on Saturday February 22, 2025 @05:06AM (#43137675)
Chat GPT series X, not compatible with Chat GPT X.

Frankly. Whoever decided on this last gen naming at MS needs to come forward. I would love to know what crazy unacceptable collection of circumstances allowed that to happen.

Re:Grok3 Launch by gulfofamerica on Friday February 21, 2025 @09:44AM (#43127991)
Off by one and naming things.
Re:Grok3 Launch by hmottestad on Tuesday February 18, 2025 @10:45AM (#43090826)
For me I was sold on the consistency. o1 does really great at several programming problems, but o1-pro does great on these problems 4 of 4 tries. I get a good answer more often with o1-pro than with just o1, or even o3-mini-high.
Re:Grok3 Launch by CamperBob2 on Tuesday February 18, 2025 @08:00PM (#43097244)
o1-pro is indeed pretty great, but I find that I can iterate several times with Gemini 2.0 Pro Experimental (or whatever their latest reasoning model is called these days) between o1-pro's responses. It's almost too slow for interactive use cases.
Re:Grok3 Launch by hmottestad on Wednesday February 19, 2025 @05:48AM (#43100730)
Yeah. I've found out that you can start out with o3-mini-high and then switch over to o1-pro, or the other way around. Helps to iterate a bit faster.