@brianchew I have a 128gb ram m4 max. If you are thinking of using the laptop for agentic coding using 30B models and above, forget it. Too slow. Also not very accurate.
@chatlyhq Scam. Their models are not what they advertise. While you may select Claude Opus 4.6, the actual model used behind the scenes is questionable.
Output generates way too fast. Level of detail is basically non-existent.
I compared this to the actual Opus 4.6 and it’s different
@Mike87_United Long term? That’s rather silly. What makes you think that anything that didn’t work in the short term would then miraculously work in the long term?
This tracks completely with what I've experienced using AI as a pair programmer. As soon as I'm tempted to let it drive, I learn nothing, retain nothing. But if I do the programming, and it does the API look-ups, explains the concepts, I learn a lot.