Fl0ps

Stacking fl0ps

[Fl0ps @exafl0ps]

@chumbawamba22 Sorry, it was in jest. Not the homeless part, im actually homeless.

[Fl0ps @exafl0ps]

@chumbawamba22 I thought QQQ leaps were quantum proof, now im homeless

[Fl0ps @exafl0ps]

@chumbawamba22 All roads lead to George Robertson?

[Fl0ps @exafl0ps]

@leveraged_cat werent you somewhat bullish memory trade a week or 2 ago? Or was that only short term?

[Fl0ps @exafl0ps]

@zephyr_z9 I mean, both of you guys were banging the drum on how insanely bullish the numbers were for memory not too long ago.

That being said, nothing wrong with changing your stance.

[Fl0ps @exafl0ps]

@zephyr_z9 What is your opinion on hynix from here on out?

The current market reaction seems quite extreme/irrational

[Fl0ps @exafl0ps]

@anni_sen Does that mean you don't think the Korean memory makers are a good play?

[Fl0ps @exafl0ps]

@chumbawamba22 Thats consensus

[Fl0ps @exafl0ps]

@chumbawamba22 faded?

[Fl0ps @exafl0ps]

@zephyr_z9 @strictrope Didn't you Samsung and hynix to trillions?

[Fl0ps @exafl0ps]

@wronguser000 $CHEX

[Fl0ps @exafl0ps]

@Evan_ss6 So then where do you hold your lit and hype?

[Fl0ps @exafl0ps]

@Wild_Randomness Where do we see this going boss?

[Fl0ps @exafl0ps]

@based16z @ContrarianCurse Definitely. I just find these benchmark comparisons foolish.

Regarding pricing power, diffusion LLMs are also interesting:

https://t.co/rOSd5PB9Wv

If these get in the neighborhood of auto regressive LLMs quality and they manage to release some open source version. Oh boy

[Fl0ps @exafl0ps]

@based16z @ContrarianCurse If you used both models you know that qwen3.5-32B is not close to sonnet 4.5

[Fl0ps @exafl0ps]

@mert I have, definitely not better than cursor atm

[Fl0ps @exafl0ps]

@Evan_ss6 aka bearish af?

[Fl0ps @exafl0ps]

@based16z What do you plan to do?

[Fl0ps @exafl0ps]

@izebel_eth You want to queue up?

[Fl0ps @exafl0ps]

@chumbawamba22 Not sure if you are still as pessimistic (or optimistic depending on how you see it) on the AGI timeline.

But I find the dwarkesh podcast with @karpathy quite insightful about the current state of AI and how much we still miss to get to the next stage

https://t.co/scEmyJv70p