i like golf and connecting dots @voltcapital | prev @Delphi_Digital
Joined June 2011
- Tweets 11,314
- Following 504
- Followers 6,106
- Likes 14,271
1,427 Photos and videos
Rob retweeted
Jun 13
local and state governments need to start treating factories like they treat stadiums
10
18
200
5,956
Most people don’t see a new tool and immediately imagine every way it could be useful. They need to see it in the context of a problem they already have.
The product can be general-purpose, but the first aha moment has to feel specific.
Relevant point given the sheer number of AI tools today.
1
5
17
2,053
in need of some brave souls willing to risk it all to make sure the super gods stay nice and helpful to humans instead of just being 'super' careful and pro bug eating
Jun 9
You're not even allowed to ask Fable about basic biology questions, let alone anything that could potentially be dangerous.
3
3
747
relevant for memory
x.com/beffjezos/status/20333…
Mar 15
Whichever is the lab that will offer continuous learning / online RL per unique agent for enterprise will absolutely print money.
Virtual headcount for all companies will become very real. Could charge $5k per month per continuous agent easily
543
less mafia, more links
Jun 10
the worst vc you know is planning to host mafia for the first time next week
1
501
science looking less like knowledge hoovering and more like experimentation throughput.
1° - fds translates domain knowledge into model search problems (seeing this now)
2° - institutions stop competing on best research start on for best experimental pipelines
3° - the great 'decoupling' where a select group of small teams w/ the right pipeline outrun incumbents.
eyes peeled for the coordination layer...
May 24
We’ve seen forward deployed engineers and AI deployment engineers.
The models are now good enough to assist with science, so we should see forward deployed scientists who can convert domain expertise into model search problems & run autonomous experiments at scale.
1
1
558
data and compute are the same bet expressed through different labor markets. ai systems are still sample inefficient, so we see them compensate with more examples more cycles.
1° - more data creates cognitive demand for cognitive labor (experts, annotators, syn operators, 24 hour feedback loops from @scale_AI, @mercor_ai, @PrismaXai)
2° - more compute creates demand for physical labor (shortages across electricians, MEPs, interconnects and all hardware @reindsummit brings people together to jam on)
3° - most ai economy outside labs is just some workaround because models still need ungodly amounts of data from the physical to workaround the unsolved learning efficiency problem
Jun 8
The AI infrastructure boom is generating strong demand for skilled blue-collar workers. In fact, there’s a shortage of electricians, fiber technicians, and mechanical tradespeople needed to build and maintain AI data centers. Meta’s new $115M America’s Workforce Academy provides paid training plus job guarantees for exactly these roles. This is the kind of practical jobs training program that we need more of.
2
2
1,271
AGENTS, CREATORS, AND THE PAY-PER-USE INTERNET
Michael Blau (@blauyourmind), founder of Drip, joins Khushi (@khushii_w) and me to talk about what happens when agents become the consumer.
We go from a forgotten HTTP error code designed for internet-native payments, to what happens when agents start paying creators directly.
01:59 What made agentic payments possible
05:24 The dev stack for building agentic commerce
11:39 AI research vs. paywalled content
16:45 Creators are still valuable, but it's about curation 20:23 From subscriptions to pay-per-use
26:25 Agents spending money on your behalf
32:19 What the agentic internet looks like
@TheDryRunPod
3
4
22
2,111
1 week till we return to the shores of the detroit river
1
2
7
1,429
the clancy-ification of our defense industrial base
🎮🚀🇺🇸 ANNOUNCING A NEW CALL FOR PAPERS ON VIDEO GAMES AND WARFARE
DARC is seeking papers addressing the intersection of video games and real-world conflict. Honoraria of $2,000 available for pieces of 2,000 - 4,000 words.
—
Video games are among the dominant cultural and cognitive substrates for the next generation of military and civilian leaders.
These leaders will have spent their formative years mastering real-time strategy games, squad-based shooters, and competitive multiplayer environments. These shared formative experiences are now diffusing through the world’s military institutions. Their narratives, reward structures, and tactical logics will inevitably shape how tomorrow’s warfighters perceive conflict, solve problems, and make decisions under pressure.
DARC believes that understanding these influences is essential to understanding how future conflicts may unfold.
In addition, or perhaps as a consequence of this upbringing, DARC is struck by how emerging battlefield platforms, from hardware to software, increasingly resemble the interfaces and control schemas of video games. The proliferation of drones, autonomous munitions, cyber warfare, and warfighter wearables has created a battlefield that, in many respects, looks remarkably like popular video games, both in terms of gameplay, environments, and story elements.
If, to turn a phrase, “the medium is the meta,” then a deep understanding of how strategic dynamics played out in the world of e-sports and competitive videogaming may give us clues to the dominant strategies of future conflict. Increasingly, it seems we may now be looking towards fighting the real-world versions of conflicts simulated decades ago.
defenseanalyses.org/video-ga…
1
1
651
Rob retweeted
May 22
Our @Kickstarter is now LIVE.
Meteors. Satellites. Drones. Aircraft. Auroras. UAP.
SkySphere combines edge AI, trusted timing systems, and real-time detection software to transform raw sky imagery into actionable, trustworthy sky intelligence.
Observe. Replay. Detect. Contribute.
Built for wonder. Designed for real-world awareness.
Back the future of continuous sky observation 👇
bit.ly/SkySphereKick_SM
3
6
29
5,183
new markets, new risk transfer tools
1° - compute cost becomes something enterprises need to price, not just pay
2° - buyers hedge future inference demand instead of eating usage volatility directly
3° - anera becomes the venue where compute costs, agent demand & margins get priced/hedged/financialized
big congrats to @intheanera. as value moves downstream of inference, the clearing layer becomes become the control point.
1
4
328
big congrats to @intheanera. as value moves downstream of inference, the clearing layer becomes become the control point.
Jun 2
We're excited to announce Polymarket has facilitated the first ever on-chain, institutional block trade in the prediction market space.
Institutions using Polymarket to hedge GPU compute exposure at scale give a glimpse into both the future & the promise of prediction markets.
2
13
914
fringe untapped data is long overdue for a rebrand
May 28
Today, we're launching shift. We're starting by cleaning your apartment in New York City, for free.
Here's how it works. Book a shift cleaning. A vetted shift operator comes to your home wearing one of our devices. They clean. They leave. You pay nothing.
In exchange, we record the cleaning. Robotics is being built on data about how people do daily tasks, and the value of that recording is what funds the service. Anything personal in it is anonymized before the recording is processed.
By now, you have heard about the shift to AI more times than you can count. About the shift toward you, the part where you actually feel it, you have heard almost nothing. Shift is what starts to make it concrete, in specific cities, with specific services.
Today, cleaning in New York. Soon, handymen, repairs, and errands across the globe. And this is just one side of shift, with more on the way.
Comment “shift” and we’ll send you an early access link.
1
1
919
been thinking about Fringe Untapped Data at @VoltCapital for a long time now
FUD is not just about collecting data. that makes you a broker
it is about designing crypto economic systems that incentivize the right collection what our AI overlords cant
- reward the edge
- verify the source
- build the loop
FUD lives where there are no APIs (eg robot arms, off grid sensors, localized human interactions). This is not public data. It is invisible until someone designs a system that makes it economically worth surfacing.
we have backed teams doing this in health, robotics and space and were looking for more in energy and defense.
moats are built on the edge
605