Built by a trader. Not advice.

Tell it what you believe. Get the tickers.

Type a belief about the world. Hunch reads the news, finds who actually gets paid, and hands back the companies and funds most exposed to it, each with catalysts, risks and what would prove you wrong.

Try it free for 30 daysNo card. Then $5/month.

GMade by Gavan, @GavanGanung

real run · Sep 16, 2026

“Humanoid robots will be doing real work in factories and warehouses within five years”

TSLAlongTesla
HYMTFlongHyundai Motor
AMZNlongAmazon.com
NVDAlongNVIDIA
JBLlongJabil

+2 more. See the full list with catalysts and risks ↓

You had the idea. The next hour is the hard part.

The usual way

✕Read article after article to work out who’s actually exposed
✕Dig through filings for the supplier nobody mentions
✕Guess whether the idea is already priced in
✕Never write down what would prove you wrong
✕Skip the hedge

With Hunch

✓Type one sentence
✓Get 7 ranked names, not just the obvious one
✓Pure plays, suppliers, incumbents and a hedge
✓Dated catalysts and specific risks for each
✓A kill switch on every list

The run below took 42 seconds.

A real run, unedited

“Humanoid robots will be doing real work in factories and warehouses within five years”

Sep 16, 202642scost 3¢ to run6/7 tickers on a US exchange

Not on a US exchange: HYMTF. Left in, because this page shows exactly what it produced.

Read before answering: forbes.com, hypernova.news, ltts.com, iiot-world.com, ibtimes.co.uk, en.sedaily.com

Humanoid robots will move from narrow caged pilots to paid full shift work in factories and warehouses within five years.

The theme is early but crowded with hype: pilots at Schaeffler Toyota Mercedes BMW BYD and Foxconn are real yet task specific. Market prices leaders for AI optionality while reliability safety and economics for full shift operation are still unproven.

Ideas · 7 · ranked by thesis fit

TSLA

long

Tesla·Pure play·years

Tesla aims to deploy Optimus for material handling and simple factory work internally then sell externally, direct exposure to humanoids doing real work.

EdgeMarket treats Optimus as an option while Tesla is already using it as internal labor to test factory tasks

Catalysts
Q4 2026 update on Optimus pilot scale and 2027 production target
Tesla Q1 and Q2 2027 earnings disclosures on Optimus units in Fremont and Texas factories
Unveil or demo of Optimus Gen 3 dexterity and cost per unit

Risks
Optimus still in controlled pilots with 2 to 4 hour battery and reliability drops outside labs
Auto margins and robotaxi expectations dominate stock and can swamp robot news

exposure common shares

HYMTF

long

Hyundai Motor·Pure play·years·adr

Owns Boston Dynamics whose electric Atlas is piloting industrial tasks at Hyundai facilities targeting phased deployment through 2028.

EdgeBoston Dynamics electric Atlas at TRL 7 to 8 is ahead of most peers but value is hidden inside Hyundai

Catalysts
Hyundai phased Atlas deployment updates through 2028
CES 2027 Boston Dynamics progress on autonomous battery swapping and 4 hour runtime
Hyundai investor day disclosure of Hyundai Mobis actuator and factory rollout plans

Risks
Atlas pilots at Hyundai facilities are still pre scale with safety and uptime unproven
Auto cycle in Korea and EV demand drive ADR more than robotics

exposure ADR shares

AMZN

long

Amazon.com·Beneficiary·years

Largest warehouse operator trialing humanoids for package transport and delivery logistics, gains if humanoids cut unit labor cost.

EdgeAmazon has brownfield warehouses built for humans where humanoids slot in without full retrofit

Catalysts
Holiday peak 2026 update on Digit and humanoid trials in fulfillment and sort centers
Q4 2026 and Q1 2027 AWS re Invent robotics and Agility disclosures
2027 capex disclosure showing automation spend per new warehouse

Risks
Deployments remain pilot scale and caged for safety with unclear total cost of ownership
Labor and regulatory scrutiny of warehouse automation

exposure common shares

NVDA

long

NVIDIA·Picks & shovels·years

Supplies the compute simulation and foundation models that let humanoids generalize beyond fixed pilots.

EdgeEvery credible humanoid needs vision language action models and simulation training where Nvidia is standard

Catalysts
GTC 2027 robotics model and Isaac platform updates
Quarterly data center plus robotics software revenue commentary
Design wins with Agility Apptronik Figure and Chinese makers disclosed

Risks
AI chip cycle and export controls matter far more than robotics revenue
Competition from Tesla Dojo and in house chips

exposure common shares

JBL

long

Jabil·Beneficiary·years

Trial partner for Apptronik Apollo and electronics manufacturer positioned to build humanoid hardware at scale.

EdgeMarket sees a contract manufacturer while Jabil is both trial site and potential scaled builder for Apollo

Catalysts
Late 2026 and early 2027 updates on Apptronik Apollo trials at Jabil facilities
Q4 2026 and Q1 2027 earnings calls for warehouse automation business
Repeat order or manufacturing contract for Apollo hardware

Risks
Apollo still in warehouse and factory logistics trials not complex assembly
Electronics manufacturing cyclicality

exposure common shares

ROK

long

Rockwell Automation·Picks & shovels·years

Provides safety controllers software and integration for brownfield plants where humanoids must work beside people and legacy systems.

EdgeHumanoids need caging safety controllers and legacy line integration before they can work full shifts

Catalysts
Q4 2026 earnings and 2027 outlook for Lifecycle Services and automation demand
Safety PLC and machine vision integration wins tied to humanoid cells
End 2026 update on Schaeffler Digit moving to uncaged operation alongside workers

Risks
Factory capex slowdown delays automation upgrades
Humanoid pilots staying narrow limits incremental content

exposure common shares

MAN

short

ManpowerGroup·Hedge·years

Short industrial temp labor expresses the downside for human labor if humanoids take repetitive material transfer and inspection support jobs.

EdgeIf 25 pound basket moves kitting and machine tending shift to humanoids, temp industrial staffing demand erodes first

Catalysts
Quarterly North America manufacturing staffing volumes and pricing
Schaeffler Toyota and BMW disclosures on pilots converting to repeat Digit Apollo Figure orders
UBTECH Liuzhou 10000 unit capacity utilization updates

Risks
Humanoids remain below TRL 9 with 60 percent real world accuracy and short battery life
Tight labor markets can lift staffing even as pilots expand

exposure common shares

Watch for

→Pilots converting to repeat orders at Schaeffler Toyota Mercedes Jabil BMW BYD NIO and Foxconn
→Digit operating uncaged beside workers by end 2026 and Atlas Optimus sustaining 8 to 20 hour coverage via swapping
→Cycle time energy use and intervention rates approaching fixed automation in material handling and machine tending
→UBTECH Liuzhou line proving sustained output toward 10000 units per year not just 10 minute design takt
→Amazon and auto makers disclosing labor hours per unit improvement tied to humanoids

Kill switch

×By 2028 fewer than a handful of sites convert pilots to repeat paid orders and most remain caged demos
×Battery endurance stays at 2 to 4 hours without reliable autonomous swapping and full shift uptime fails
×Total cost of ownership stays above human plus fixed automation on cycle time and intervention rate
×UBTECH 10000 unit plant and Chinese unit counts do not translate into sustained commercial deployments

What it’s good at

✓Finding second-order names you wouldn’t have thought of
✓A structured first pass in about a minute
✓Making you look at the hedge and the exit before you’re in love with the idea

What it’s not for

✕Price targets, entries or timing
✕Knowing your portfolio, taxes or risk tolerance
✕Being right. It reads the web, so it can be wrong. Check before you act.

Hey, it’s Gavan

@GavanGanung

I trade my own account and build my own tools. Having an idea was never the slow part. The slow part was the hour after: who actually gets paid, what’s the catalyst, and what would make me wrong.

Hunch does that first hour in about a minute. It’s a first pass, not a verdict. It reads the web, so it can be wrong, which is why every list ends with a kill switch.

A run costs a few cents to operate. Five dollars a month covers that and keeps it simple: no ads, no brokerage deals, no premium picks upsell.

Not advice. One guy trading his own account.

Open numbers

Straight from the database, updated every time this page loads.

People who tried it: 1
Hunches run: 1
Paying: 0
MRR: $0
AI bill so far: $0.04

One plan

$5/month

30 days free. No card to start.

✓10 hunches a day
✓Live news research on every run
✓Catalysts, risks, a hedge and a kill switch
✓Your history, saved
✓Cancel from your account page

Try it free for 30 days

Questions

Is this investment advice?+

No. It’s a research starting point. It doesn’t know your situation and it can be wrong. Treat every name as something to look into, not something to buy.

Where do the ideas come from?+

Each run searches current news on the web, then an AI model (Meta’s Muse Spark) works out who gets paid and ranks the ways to express your belief. You can watch which sites it reads while it works.

Why only $5?+

A run costs a few cents to operate. The example above cost 3¢. I’d rather a lot of people use it than a few people pay a lot.

What happens after the 30 days?+

Nothing, unless you subscribe. There’s no card on file, so there’s no surprise charge.

Can I cancel?+

Yes, from your account page. You keep access until the end of the month you paid for.

What do you do with my hunches?+

They’re saved so you can come back to them. They aren’t sold to anyone.