$TASK

TaskUs Inc Class A

Share on X

$TASK

TaskUs Inc Class A

Past 12mo:3mentions1voices

Share on X

All voices3 mentions since May 2026calls only

Loading price chart…

Frequently asked

Who mentions $TASK most on Surfaced?

@crux_capital_ (Gaetano) mentions $TASK most — 3 mentions in the past 12 months.

When was $TASK first mentioned on Surfaced?

The earliest tracked mention of $TASK in the Surfaced archive was on May 29, 2026.

How many voices track $TASK?

1 FinTwit voice has mentioned $TASK 3 times in the past 12 months.

What kind of voices mention $TASK most?

$TASK is mentioned most by Semis voices on Surfaced.

Recent mentions

3 mentions

@crux_capital_Jun 4, 2026 · 12:00 PM

When I talk about Physical AI I am talking about so much more than Humanoids. In fact, I am not even thinking about humanoids a majority of the time. Layer 1: Training and Model Development Companies like $NBIS $U $PL $BKSY $INOD $TASK etc. Layer 2: Simulation and Synthetic Data Companies like $BBAI $CAE $U $BSY $SVCO etc. There are 9 more layers. My goal with my Physical AI thematic coverage is: > Educate on all the layers > Introduce the best companies for each layer > Strategically invest + share weightings That's what coverage looks like on Substack. > In depth education > Unpacking the companies > A live portfolio as I build it out

Open on X

@crux_capital_Jun 1, 2026 · 10:56 PM

Layer 1: Training and Model Development $TASK $NBIS $INOD $U $BKSY $PL etc. Before any Physical AI system can work in the real world, it needs to learn. That sounds obvious, but I think it is a really important place to start because is is very easy to jump right to the machine itself. Whether it is a robot, a drone, an autonomous vehicle. The humanoid. Whatever visual example is easiest to picture and what people think about when they hear 'Physical AI'. But before any of that becomes useful, there has to be a learning process behind it. A physical system needs to understand: > what it is seeing > what it is trying to do > what a good outcome looks like > what a mistake looks like > what action should come next. That can come from: > real-world data > synthetic data > video > simulation > human demonstrations > expert feedback etc. There are so many exciting companies in this layer and it is often overlooked when we think about Physical AI thematic

Open on X

@crux_capital_May 29, 2026 · 1:33 AM

Tomorrow I am presenting my first investment In the Physical AI thematic My process is this: 1) Research the layer 2) Research the companies 3) Present the company thesis + weight for Physical AI portfolio Rinse and repeat until I get through all the layers (11) By the end, I will have a fully built out Physical AI portfolio that I believe is best positioned for the next few years Here are some of the companies I screened for the first layer: $PL $BSKY $U $INOD $TASK $NBIS $SPIR $REKR and more If you want to join for the full research and buildout of this portfolio, click here and scroll to sub! https://t.co/ubIy5CKh9Y

Open on X