i just made my first ever @Dune dashboard, on a subject i've been digging into a lot lately: stock tokens on robinhood chain.
it covers daily mint/burn stats, dex volume, holder breakdowns and which protocols/infra actually hold these tokens, and where supply sits across the different stock tokens, which ones carry the most supply, which ones trade the most.
if you're into this stock tokens meta, could be a useful one to dig through.
open to feedback if you know your way around dune dashboards, will keep updating it as i go.
dune.com/skobro/rh-stock-tok…
and here we go again, it keeps going
to be clear it’s an offer not 60k layoffs. but eligible is ~27% of the workforce. in april it was ~7% of us staff, the first buyout in microsoft’s history. so in 6 months the scale went from 8k to 60k
the ai link is mostly about money. microsoft spent $37.5b on data centers in one quarter, so payroll has to shrink somewhere. 15k jobs cut in 2025, a buyout in april, now this. meta cut 10%, amazon 30k, block 40%
the second phase is clearly started
BREAKING: Microsoft, $MSFT, is offering buyouts to 60,000 workers as part of the company’s efforts to cut costs amid AI adoption, per Bloomberg.
Personnel on select teams in the US, Canada, India, and the UK at a vice president level and below will be eligible.
This reflects ~27% of Microsoft's total workforce.
not sure i buy the abundance era story where work becomes optional and we all get a weekly check
but one thing i'm sure of. ai already widens what's possible and it keeps going
people always had ideas and never had the means to ship them. now if you want it and put in the minimum effort you can build, iterate and ship almost anything that crosses your mind
it amplifies values i think most of us hold dear. freedom, creativity, agency
and it makes me really excited for what's next
i went through the latest data from the biggest sources on ai and jobs: wef, ilo, imf, mckinsey, goldman sachs, anthropic, yale
put everything in one visual
left: sectors from most to least exposed
right: what each source actually measures (potential, projections, employer intentions, observed effects)
tldr
most exposed: admin, data entry, customer service, banking and insurance
least exposed: agriculture, healthcare, hospitality
thoughts on robotics
when most people think about this sector they picture the domestic robot. the one doing chores at home for a family
i think the real value is somewhere else. robots in warehouses and factories doing manual repetitive work in buildings already built around human bodies so you don't rebuild the building, you just change the worker. one that could be cheaper and work around the clock.
the question is who already has robots with real proof of work and traction.
agility is one i'm watching. digit moved 100k+ totes at gxo with ~98% accuracy while on task. deployed or committed at 9 customer sites incl schaeffler, toyota canada, amazon. and it's going public through a spac.
but the scale is tiny. ~150 units shipped in 2025
unitree is the opposite. 5,500+ humanoids shipped in 2025, the most in the world. but most of the revenue seems to come from research and education, only a small share from factories
ubtech is the chinese one aiming at factories. 800m yuan in orders and several hundred walker s2 delivered
then optimus, the most famous one. it starts in tesla's own factories, no outside customer yet
next step for ai is getting a body so it can do physical work. that's why i think this sector is worth watching closely. my bet is the real shift comes when several customers reorder at scale
brent closed friday at $116.72, its highest close since april 29. wti, the us benchmark, finished at $103.27
the trigger was a new wave of houthi attacks on saudi arabia, energy targets included. when aramco told staff to leave its dhahran hq on thursday, the first full evacuation of this conflict, the market priced in the worst
i'm watching whether the strikes move from airports to production sites, pipelines or export terminals, because the east-west pipeline is the workaround everyone leans on. then shipping through hormuz and bab el-mandeb, and what the fed and big tech earnings say next week about oil feeding inflation
wow, seems like moody’s is more optimistic than s&p and fitch.
probably waiting to see how the budget negotiations play out
french 10y oat just went above 5% for the first time since 2002
the spread over bunds is the widest since 2011
and tomorrow moody's reviews france, currently aa3 with a negative outlook
one notch down would match what s&p and fitch already have, two notches would put moody's below both
any prediction for tomorrow ?
with ai finding more exploits every day, assets and programs are getting compromised left and right, while btc has been tested for years like the golden asset to crack. so far the base protocol has held, and i can imagine more people converging toward that thought and reinforcing the conviction
on the other side, gold vs btc as a store of value has always been a debate. marathon fusion's announcement going viral is forming a bias around the idea that nobody can make more btc while gold can supposedly be made with fusion. it's still negligible, but the idea is out there
interesting to have these two things happening at the same time, both leaning toward a bullish btc bias
non exhaustive sum up of what the cyber community now calls black october
php. two critical flaws that let attackers run code on servers without logging in, found with ai and patched oct 21. the netherlands, finland and belgium took public sites offline
certificates. 8 certs with keys so weak one was literally the number 4, about 30,000 to replace
cairn. an ai agent toolkit tested 1,100+ attacks, deere and cummins among the victims
railway and zeabur. customer secrets exposed, zeabur is shutting down in december
litellm. a poisoned update stole ai provider keys
signal. a call bug that could keep sending audio after hanging up
ahold delhaize. systems offline for over a week
meta's muse. 6,800+ private conversations exposed through shareable links
are we cooked chat
some context on barclays. 5,000 jobs is about 5% of its headcount, around 60% of it in support functions including developers, and 1,000 were already cut in the first half
it's not distress either, q3 profit rose 12% to £2.3b, though expenses came in higher than expected earlier this year
the forecast behind it: morgan stanley says european banks could cut 10 to 20% of jobs over five years, up to roughly 400,000 roles, double its january estimate, with much of it through voluntary exits and retirements
hsbc is reportedly looking at around 20,000 and standard chartered at close to 8,000 support roles
so barclays looks like one of the first of many
BREAKING: Barclays cutting 5,000 jobs in AI-led efficiency drive
trib.al/PslgiEk
10x cern's record sounds big but with no numbers published it's likely still picograms
alice made about 29 picograms in four years, so even 10x is tiny
between that and tonnes of gold a year you need commercial fusion plants that don't exist yet, mercury-198 which is only about 10% of natural mercury, and 15 to 18 years of storage for the radioactivity
but ig overall it is bullish for btc
Last month, we demonstrated high-rate alchemy: turning mercury into real, stable gold, beating CERN’s record by a factor of 10.
Last week, we enriched isotopes in the world’s most powerful plasma centrifuge, beating the Soviet record that stood for 40 years.
Now, we’re building a fusion reactor in San Francisco to achieve record power density.
french 10y oat just went above 5% for the first time since 2002
the spread over bunds is the widest since 2011
and tomorrow moody's reviews france, currently aa3 with a negative outlook
one notch down would match what s&p and fitch already have, two notches would put moody's below both
any prediction for tomorrow ?
if spacex and tesla merge, the interesting part is that the synergies are already on paper
chips. they're building terafab together in texas, $16.8b to start, to make chips for optimus, cybercabs and spacex's planned orbital data centers.
energy. tesla has sold about $890m of vehicles and batteries to spacex and xai since 2023, mostly megapacks for ai data centers. spacex's ai compute capacity reached 1.4 gw by the end of q2 and all of it needs power
ai and robotics. musk points to shared work on ai software and robotics, and tesla already put $2b into xai, which became a small stake in spacex
connectivity is the fourth area he cites, with starlink on one side and cars and robots on the other
right now all of this runs through related party deals and a terafab arrangement that is reportedly only a general framework. a merger would turn it into one company
my read is that it's less rockets plus cars and more one company owning the chips, power, compute and machines of ai
two critical php flaws sat there for 14 years and ai found them in days
remote code execution without even logging in, on software behind nearly 70% of websites
found by calif and wiz with gpt-5.6 sol and gemini 4 argon
sol isn't even openai's best model, gpt-6 astra has been out since september and i'd guess stronger ones aren't public yet
bot, muse, dots, duet
spacexai, meta, openai and now anthropic all have a personal agent
just joined the waitlist, looking forward to trying this
$NFLX is down about 45% in a year and q3 wasn't even bad
12.9b of revenue, 3.66b in profit, 325m subs
but the slowest revenue growth in 3 years, watch time up only 2% in h1 and emmy nominations at a 10 year low
and per nielsen it's 7.8% of us tv time vs 14% for youtube
profit isn't the problem, growth and engagement are
they just hired peter rice, the ex disney tv exec known for modern family and the bear
can he turn it around and bring a fresh start
every skobro pose so far
some were made with different llms, so textures, colors and details don't fully match, and i'm thinking about redoing a few to make the set more consistent
i also want to create new poses
and i split the account into 2 formats, one of them is shorter videos where skobro doesn't appear
what poses should i add next
midterms are on november 3 and polymarket has a democratic sweep at 65%
on one side, a democratic congress could push back on the iran war, which could ease oil and yields
on the other, maybe it also brings more pushback on ai and data centers, the trade holding up the market
one risk could fade while another one builds, not sure which one matters more
if energy becomes the currency, there are 3 places to stand
produce it
deliver it
turn it into intelligence
right now the scarce one is the second
→ data center power demand expected to rise 170% by 2030
→ us interconnection queue around 2,600 gw of projects, more than double the existing grid
→ median wait from application to operation: about 5 years
data centers get built in months, grid connections take years
i think the currency isn't energy itself, it's the connection to it

















