remove/noise

If you lose say nothing,
if you win say less.

[remove/noise @removethenoise]

@BlueCrystal1979 update?

[Xiaoyin Qu @quxiaoyin]

Turns out Elon is right again. The shittiest layer in AI is the model layer. The real money in AI is in compute, energy, and applications. Elon has all 3. Without Chinese open weight models, OpenAI and anthropic would have been happy oligopolies and made trillions. Now, they hired the best talents, burned billions and built the newest model, only to have Chinese free models wiping out all your margins. Every other layer is making money except for you, even though you invented the whole thing. That must feel shitty as hell.

[remove/noise @removethenoise]

@MaxKongerskov @kimmonismus Clown comment. No one cares for it to beat nvidia out of infancy. Etched got top cream talent team, look them up. Game is changing, TSM rarely invests in inference but invested in them. So they’ll get allocation of hardware from TSM to scale.

[remove/noise @removethenoise]

@MoodyWriter13 @italianinvesto Why anti heat pumps, what’s wrong with them? Mine is one of the best comfort upgrades I ever made, cop 1:6 too

[remove/noise @removethenoise]

@rosyprosperity ah my bad, misunderstood

[remove/noise @removethenoise]

@rosyprosperity "And what happened when it did not work" ?

[remove/noise @removethenoise]

@derekquick1 If this is parabolic…

[remove/noise @removethenoise]

@ackmeni They just buy from china. They are deeply depended on china. China is the bottleneck in this discussion, not axti.

[Moody @MoodyWriter13]

As a retail investor juggling a career and family, following people on X who do this full-time and push out a constant stream of supposedly high-return ideas and companies, within an environment where seconds can decide price moves, is a game with clearly uneven odds.

It’s roughly like entering a world championship of ass-kicking on one leg: there’s little to win, but a high likelihood of taking a beating.

From experience, I can say: life is a lot more pleasant when you choose not to play that game.

In the end, a mismatched investment style is just as stressful as a mismatched lifestyle.

One way to sidestep this entirely is to pay for a good Substack. In doing so, you’re essentially buying time and reducing noise-related stress.

I can say this without bias, as mine is currently still free.

[remove/noise @removethenoise]

@StockSavvyShay @Speculator_io Insane how all these lists miss AOSL

[remove/noise @removethenoise]

@SuperDuperInvst @growthrapidly AOSL is underrated af

[remove/noise @removethenoise]

@stonkmane69 @ParadisLabs I agree on everything, but all that is priced in. NVIDIA’s strategy prevents any single vendor from capturing outsized rent.
AOSL’s value is valuation compression + net-cash balance sheet, not superior rent capture.

[remove/noise @removethenoise]

@stonkmane69 @ParadisLabs How so

[remove/noise @removethenoise]

@stonkmane69 @ParadisLabs The bear case is lazy. AOSL is the best risk/reward in the entire 800V power semi group right now

[remove/noise @removethenoise]

@stonkmane69 @ParadisLabs On what metric?

Value: AOSL is the cheapest (1.0–1.5x P/B, 18–22x forward P/E).
Balancesht: Only net-cash name with 0 debt.
Execution: Real revenue + shipping products + named NVIDIA partner.
Torque: Highest beta to the 800V wave that TXN and NVIDIA are flagging as multi-year.