@FlippedfaceJ

Ambassador @Prismaxai esteemed @Arcium Trader @hypernovax

bangkok, thailand
Joined September 2021
Honored to be part of the first @PrismaXai Ambassador Cohort and proud to represent thailand 🇹🇭 community. Being selected from over 1,550 applicants means a lot to me, but more importantly, it comes with greater responsibility moving forward. My goal is to help onboard new members from around the world especially from thailand and introduce more people to PrismaX and its vision. This is only the beginning of the Physical AI and Teleoperation era. I’m excited to contribute by helping more people understand PrismaX, recognize the value of this technology, and gain clearer access to its products, opportunities and ecosystem. Thank you to everyone for the support. I look forward to growing alongside both thai community & the global PrismaX community🦾👀🇹🇭
From 1,550+ applications worldwide, 14 ambassadors have been selected for the first PrismaX Ambassador Cohort. United by a shared belief in the future of physical AI, they will help grow the PrismaX community around the world. Meet the cohort 👇
10
33
818
flippedface(ARX MODE)☂️ retweeted
RoboCon is coming to SF Tech Week. This Friday, PrismaX and @moonlake host the RoboCon Physical AI Salon: one panel, live robot demos, and a room full of people getting robots out of the lab and into the real world. @Techweek_ #SFTechWeek
19
16
5
86
3,981
flippedface(ARX MODE)☂️ retweeted
Day one at IROS, and booth 1122 is open! Teams training VLAs, world models, and robotics foundation models are stopping by to see our demo and talk through what they're building. #IROS2026 @ieeeiros
31
25
12
141
5,573
flippedface(ARX MODE)☂️ retweeted
Our monthly Office Hours ran last week. Two things carried the session: Robotics Data going live and IROS this week. Here's the recap 🦾
30
26
11
151
6,407
flippedface(ARX MODE)☂️ retweeted
Want to trade and try PopDEX before Open Beta ? Now’s a good time. I’m launching a trading competition for my referrals until PopDEX Open Beta goes live. If you trade through my referral during this period, there’s a win win opportunity for everyone. Reward 1: I’ll be giving surprise cash rewards to some of my top traders after Open Beta launches. Reward 2: You can also participate in the PopDEX Epoch 5 rewards through your trading activity. I’ll be tracking activity through my referral dashboard and rewarding some of the top traders from my side. Trade, share your experience, post your trades, and tag me so I can follow your journey. Open Beta is getting close. Join Now - app.popdex.xyz/tools/app-lau… Get on the board. Don’t sleep on PopDEX.
6
2
15
352
PrismaX just opened robotics data in limited release. What I found interesting is that they’re not just focused on collecting more data. Small details like camera placement, how objects are arranged, and how an operator performs a task can all affect what the model actually learns. PrismaX also trains on the data they collect and uses the results to improve how they collect the next round of data. The picture is getting clearer: VLA Foundry → people collecting & reviewing data Robotics Data → data for teams building Physical ai PrismaX started with teleoperation, but it looks like they’re moving beyond that now
Usable, diverse, proven demonstrations for the teams building physical AI. Now in limited release.
2
5
116
$btc just tapped $87k zone and is taking a breather, but solana:So11111111111111111111111111111111111111112 is still showing pretty solid relative strength. my entry around $116 came from a pretty simple setup: $112–116 held up well, sol reclaimed $110, and price structure kept printing higher lows. now price is back near the $120 resistance! if sol can break this zone with volume behind it, things could get a lot more interesting. if u guys looking for perp that actually gives value back to traders, you dont wanna sleep on @popdex_ clob, fast execution, markets keep expanding including rwa perps and reward system that's not just about chasing volume. closed beta is heading into its final stretch. and don't miss the biggest reward of the final round that's coming up october getting close... might be cooking up something 👀
2
52
I have farmed dozens of protocols waiting for institutions to show up, Canton is finally the network where they stand. While @TitanExchange_ is the venue that brings privacy execution to perps. Just claimed my pass, here is a link to the waitlist waitlist.titanx.cc/?ref=2D1F…
1
38
flippedface(ARX MODE)☂️ retweeted
Notice anything different?
72
33
8
225
11,536
flippedface(ARX MODE)☂️ retweeted
No robot is safe. Not the one dancing, not the one folding laundry. The tolerance for a wrong move is zero. You can't sandbox a 50 kg machine or roll it back. An LLM that gets it wrong gets regenerated. An arm that gets it wrong goes through the table.
For those who don’t know, these robot moves are essentially remote controlled. The robot rather blindly performs those while maintaining balance and such. AI safety for physical robots makes the current crop of AI safety for large models look like child’s play. Extremely important to get it right.
14
7
7
99
7,165
AI safety for physical AI is different. for LLMs, a wrong answer can be bad. for robots, a small error can damage hardware, break parts, or even hurt someone. the biggest challenge isn’t just preventing bad behavior, but minimizing physically inappropriate actions as close to zero as possible. sometimes, the safest training data is simply the data you choose not to put into the model in the first place. that’s why data quality and validation matter so much for physical AI. and this is one of the areas @PrismaXai is working on. 🦾
No robot is safe. A 50 kg machine on lithium batteries with several horsepower of peak output is heavy machinery, AI or not. That makes safety for physical AI easier to reason about, not harder. Full piece 👇
2
83
flippedface(ARX MODE)☂️ retweeted
Our cofounder and CEO @castorhat is speaking at IROS 2026. What Makes Good Robotics Data? Tech Talk Stage, David L. Lawrence Convention Center Tuesday, September 29, 3:15 PM His answer starts the same way every time: it depends on what you are training 🤔
27
13
13
108
9,786
flippedface(ARX MODE)☂️ retweeted
AI safety started as a commercial problem for LLMs. Robots inherit it, plus physics. We map each LLM safety failure onto physical AI.
21
16
28
112
17,268
flippedface(ARX MODE)☂️ retweeted
IROS 2026 is two weeks out. Find PrismaX at Booth 1122 in Pittsburgh! PrismaX is the service layer for physical AI: training data collected to your spec, scored to a standard, delivered in the formats your pipeline reads. More on what we do below ⤵️ @ieeeiros #IROS2026
32
22
24
144
12,817
flippedface(ARX MODE)☂️ retweeted
Humanoid installations went from about 2,000 in 2024 to 15,000 in 2025. Barclays projects 60,000 this year. Every one of them is a data collection tool. That is the part of the number that matters.
24
16
31
141
11,881
flippedface(ARX MODE)☂️ retweeted
Different environments, one collection protocol. Each task is specified up front: objective, action type, variations. Each episode records three camera views and is scored to our standards before it trains a model. Varied on purpose: scene, objects, lighting, handedness.
28
16
12
114
7,671
flippedface(ARX MODE)☂️ retweeted
The simplest motor usually wins. That's the whole argument in most hardware right now. Austin Brown, who designs motors at @BostonDynamics, said it plainly on Doesn't Grasp episode 2 with @castorhat.
21
12
11
120
5,703
flippedface(ARX MODE)☂️ retweeted
Validator Watch Wednesday is here! This week's reviews got into some genuinely interesting details 👀
24
12
10
137
6,497
flippedface(ARX MODE)☂️ retweeted
Cards are coming to Fast. We’re partnering with @PulsarMoneyApp to bring card payments into the Fast ecosystem. From intent to purchase, without breaking the flow.
19
21
6
113
9,886