Tom | LD

Head of Research at LD Capital | Co-Founder of a UK banking startup | Former Vice President at Morgan Stanley in London

[Tom | LD @tom_ldvc]

Had a great conversation covering some of the biggest questions in crypto today.

We discussed where the market could be heading over the next few years, which sectors I believe have the strongest long-term potential, and how we think about capital allocation in the current environment.

I also shared how we approach investing at LD Capital - from sourcing opportunities and evaluating teams to supporting founders after we invest. It was a great chance to explain our investment philosophy and what we look for in projects that can create lasting value.

Thanks for the great discussion 🙌

[Paul Grewal @iampaulgrewal]

Yep. Nobody ever gets everything they want with legislation. But they can get what they ultimately need. It’s time for Congress to close out the last issues, pass this bill and finish the job.

[Michael Nadeau | The DeFi Report @JustDeauIt]

The two-year (which tends to lead the FFR) is now 56bps higher than the upper range of the Fed Funds Rate.

Meanwhile:

- Round 2 of the hot war in Iran is now escalating
- The SoH is effectively closed for commercial shipping
- Ukraine has knocked out 40% of Russia's refining capacity
- WTI is up 30% over the last few weeks

And Warsh made the following statement in the June presser:

"We have the capability and commitment to deliver on our price-stability objective of 2 percent. That’s exactly what we’re going to do.”

It's all happening as Tariff Wars reignite (Canada) and a hotly contested midterm election season shifts into gear.

----

We shared a full "Cycle Awareness" update for BTC today in which we assess whether the onchain data + macro setup points to whether the "cycle low" is in or not (and how we're playing it).

If you'd like to check out the latest research, see the link below 👇

[Imran @lmrankhan]

I’ve been using Tesla FSD for over a year and completely take it for granted now.

I recently rented a car on a trip and felt like I had to relearn how to drive. FSD has gotten so good that I can’t imagine buying or even renting a car without it.

I can see a world where the only deciding factor when choosing a car will come down to whether it has FSD. If it doesn’t it’s not even on the list.

[IcoBeast.eth🦇🔊 @icobeast]

I had 0 marketing/GTM/UA/Product experience prior to joining Kalshi - I'm classically trained as an engineer and attorney.

But I really like winning and will always find a way to do so.

I wrote our playbook for growing crypto predictions.

I want to hire people who have that same drive (dog in them) to help me grow our non-crypto products at Kalshi.

If you can do this in 8 months for our commodities markets, upcoming perps offerings, or any other product/market category we have on the exchange that you're passionate about, get in touch.

[Brian Armstrong @brian_armstrong]

If regulators can secretly debank an entire lawful industry, no business is safe.

That’s exactly what nearly happened to crypto, and we exposed the details.

The Gensler SEC deleted texts at the height of the anti-crypto campaign, FDIC buried evidence - it was all uncovered after we fought to expose the truth.

This win is not only for us, but for every American, and every American company expecting transparency and accountability from the government.

[Nic @puckrin]

This chart should alarm every US frontier AI lab.

In January 2025, US models dominated global usage.

By April 2026, China had taken the lead.

It hasn't looked back since.

Chinese models now run nearly 60% of all AI token traffic on OpenRouter.

[Imran @lmrankhan]

Heard the unfortunate news about Network School.

I’m not sure moving countries solves the underlying issue. It changes your sovereign dependency but doesn’t remove it, and for Balaji’s vision to come to life, it needs to own everything end to end.

From a software analogy perspective, you need to own the full stack: land, laws, infrastructure, economy, and citizen experience. Otherwise, you’re an app running on someone else’s operating system, and they can change the rules or shut you down.

I get that it’s a chicken and egg problem get users and citizens, build a community temporarily, then reestablish elsewhere once you have momentum. But you’re not competing with typical competitors you’re competing with nation states.

And your users aren’t just dealing with a specific niche problem you’re solving, like shopping or payments. It’s all of it, from things like healthcare to where citizens stay, how they transport themselves, and more.

So if you get too powerful, nation states will cut you off. Getting users acclimated to a new city is also extremely hard and will likely face adverse selection. Founders who want to get to PMF quickly will tap into Silicon Valley, NYC, or other global cities with established network effects.

Praxis showed how hard this is. They built the network and raised capital but still struggled to secure land, legal autonomy, funding, and enough residents to build the actual city.

I do believe in what Balaji is doing, and it is the ultimate fail safe plan. But I think it requires an Elon Musk level, Herculean effort, with large corporate partners or billions of dollars in capital to get it off the ground.

[vitalik.eth @VitalikButerin]

A new type of "high-level programming language" that seems really worth trying to make, is a language that gets compiled to Lean (or HOL, or...) that is specifically about making it as friendly as possible for a human to read definitions and theorems.

Not the proofs - as all that matters with proofs is that the proofs are correct - just the definitions and theorems.

The intended use case is that AI outputs a blob of proofs, and you're trying to make it as easy as possible for anyone reading the output to understand what the actual precise claims are that have been proven

[DCinvestor @DCinvestor]

i don't think Ethereum is commie (though there are commies who like it), but real talk:

Ethereum's branding over the past 5-10 years has held it back and made it unattractive to many audiences who SHOULD use it

an overt focus on "magic and whimsy" (by the EF and others) rather than on hardened public architecture harmed the public perception (and possibly adoption) of Ethereum

which is a shame because in many ways the core of Ethereum is much harder than anything else in programmable crypto

but there are now new institutional players in the space though with serious exposure to Ethereum who will want to appeal to those audiences directly

it's time to shift the focus from "unicorns and rainbows" to "security which exceeds the guarantees provided by even nation-states"

seeing signs of that shift already, even at the EF

[Brian Armstrong @brian_armstrong]

Capitalism is a powerful force for good. It benefits everyone, especially those who lean in and help build the future.

But it’s under attack (again). Wealth taxes, price controls, government-run grocery stores, asset seizures. Fear is being used as a political lever and an excuse for control. I’ve seen this play out before, it leads nowhere good.

I lived in Buenos Aires in 2008 when the Argentine government seized $30B in private pensions to ‘protect’ citizens from the financial crisis. Then they froze access to the dollar and nationalized companies. Within a decade, inflation was over 50%. The elites had miraculously moved their money, leaving regular people to take the financial hit.

Politicians are selling collectivism out of the same playbook used by Argentina, Venezuela, and the former Soviet Union: promise protection, concentrate power, tighten the fist. And the costs fall on the poorest people in society who are unable to protect themselves.

Capitalism isn't perfect. But it’s far better than every alternative that’s been tried. Millions of people making independent decisions, with the free market rewarding the best ideas, will out-build any central planner.

[Market Bubble @MarketBubble]

Jesse Pollak on the ecosystem forming around tokenized inference

"There's this really cool tokenized inference market that Venice has innovated, where they've taken inference and turned it into a token called DIEM where you can buy it and you get $1 of inference per day"

"That innovation of tokenizing inference has unlocked a ton of innovation around it, where people are using that tokenized inference to do things"

"There's other projects that let people sell their tokenized inference at a discount so other people can get cheaper inference and the whole system works more efficiently"

"I feel like that's the first time we're really seeing the intersection of AI plus markets plus crypto and it's thanks to Venice and it's happening on Base"

[IcoBeast.eth🦇🔊 @icobeast]

Crypto is the only industry in the world where you can make $20 billion dollars and still be told you're stupid by dudes fighting for $17.31 in the shitcoin trenches because you didn't support their coin lmao

[Archer Labs @degenarcher28]

Cobie comes up with an idea that will resonate with the CT crowd. “This will definitely improve Base’s image” he thinks to himself. In order to put his idea in motion, he books 46 calls with the legal team at Coinbase. After 2 weeks of intense discussions, his marketing plan is approved by the legal team. Although he has to compromise quite a bit to make sure it can pass. His plan is now compliant.

Now the difficult part. He has to present his idea to Brian since it includes posts from Brian’s twitter account. He books an in-person meeting with Brian and the other Coinbase executives. He has to wait 24 days for their availability.

During the meeting he has to convince 8 people that are disconnected from reality about his marketing plan. The competition is fierce. Corporate politics mean the other people in the meeting see him as a threat to their existence so they are inclined to refuse whatever he puts out and agree with whatever Brian will say.

During the meeting, everyone outside of Cobie is very much surprised. “But wait these are real people you want to appeal? I don’t get it.” Brian says in a visibly upset tone, hinting at this meeting being a waste of their time. Rest of the executives agree with Brian. Meeting ends and @cobie is told there will be a decision in two weeks.

Two weeks go by, his idea is refused.

He walks around the office, confused. While walking back to his corner office, he sees a room where an intern is actively being tortured. The intern opens up Jesse’s twitter account and tweets:

“it’s a good day to build on @base”

[Nic @puckrin]

The US Petroleum reserve just keeps collapsing.

It fell 5.1m barrels last week and now sits 311.4m.

This is 35m barrels lower than the 2022 low.

It's the lowest level since 1983.

It's 43% of the authorised capacity.

This will take years to refill.

[Jim Cramer @jimcramer]

We must NOT let our companies use these Chinese models to save a few bucks. OpenAI and Anthropic are correct. This is vital national security. Please read Bing West's just released Cat 5. I respect the Chinese people greatly but these companies are run by the PLA for heaven's sakes.

[Brian Armstrong @brian_armstrong]

Regarding the recent buzz around my profile picture changes and some feelings that the Base community isn’t being supported enough: I appreciate the feedback (even if tough to hear) and I realize you wouldn't take the time to respond unless you cared.

It seems I wasn't clear enough setting expectations for all of you, so let me try and do that now: please don’t follow my personal X account for investment advice or signals around individual coins. I’m simply posting things I find interesting or funny on the internet. I may not even be aware if there is a coin or project attached to the content I'm posting (you should assume I'm not, to be safe). My posts and profile pics are also not endorsements or commitments to anything.

Base is our shot at building foundational infrastructure for financial services where we can all innovate together. That includes tokenized stocks, borrow/lend protocols, stablecoin payments, etc and yes even meme coin trading. I believe in economic freedom, and support you trading whatever you want. Just realize that if you're treating my X account as alpha, you are doing so at your own risk, against my wishes. I would never recommend this.

On the topic of “support” from the Base team or Coinbase, there seems to be confusion around this too. There are many tokens we would like to list on our centralized exchange, but can't for compliance and regulatory reasons. If you want Jesse or me to pump your bags or shill certain coins, we're also not going to do that.

Here is what we are committed to on support: we run in-person Base Batches and issue grants to high potential builders, Coinbase Ventures and the Base Ecosystem Fund also invest in the most promising companies building on Base, and we periodically integrate promising Base defi protocols into our products at Coinbase which enhances distribution. I might even mention something on X if I think it's interesting, but again, this is not investment advice. In general, we bias toward supporting projects that we believe will create long term customer value.

You may like this approach or you may not, but either way hopefully it creates clarity going forward, so you can decide what you want to do. I respect your choice, either way.

P.S. I still might post memes if I think they are funny! It's still not investment advice.

[Bob Elliott @BobEUnlimited]

Only 2% of investors think that a hard landing is possible. Lowest in years.

54% of investors think we are on a growth rocket ship ahead. Highest in years.

[Michael Fridman @justmikeel]

5 TMT events/takeaways from this past week:

1/
SOX -10% on the week, worst in 15+ months, bear market territory (-20%+ from June high). Moonshot AI's Kimi K3 (2.8T parameter open-weight model) certainly not helping. But the move was a leveraged de-grossing, not a re-rating. The hardware trade was the single most crowded long in global equities. GS High Beta Momentum is now -33% from its June highs after briefly touching +60%+ YTD three weeks ago. Outstanding results (TSM record quarter, ASML beat) produced selloffs. That is not what fundamental deterioration looks like. That is what peak-expectations + maximum crowding looks like when a narrative shock hits. $NVDA $AMD $MU $INTC $AVGO

2/
The Kimi K3 question that actually matters isn't the benchmark, it's whether the "age of scaling" is giving way to the "age of research." Moonshot AI, unable to match the largest Western pre-training runs, closed most of the capability gap through architecture, RL, and post-training. If research techniques keep compounding at this rate, the marginal return on another dollar of pre-training compute keeps falling. That is the question stress-testing the hyperscaler capex thesis.
The bear case has a data problem. Yipit Data's latest token pricing analysis shows effective AI token pricing down only 6% YTD, a stark contrast to the -39% collapse in 2H25 when xAI, https://t.co/VWWteZjyih, and Google Gemini 2.0/2.5 aggressively took share. The market is stabilizing, particularly among enterprise buyers who remain willing to pay for leading model performance. OpenRouter volume has gone from effectively zero to ~4-5x cloud volume in roughly six months. Total realized token price has been essentially flat since January 2026 while overall token consumption has grown 5-6x. Lower-cost open-source models are expanding inference TAM, not cannibalizing it. That is the chart the Kimi K3 bears need to explain. See chart below.

3/
$TSM reported its fifth consecutive record quarter: $40.2B revenue (+36% YoY), net income +77.4% YoY, gross margin 67.7%, CapEx raised to $60-64B. Stock sold off -2.3%. Same dynamic as $ASML the prior week. The market had anchored to earnings as the bullish catalyst. Once the results arrived, attention immediately shifted to 2H margin dilution from the 2nm ramp and US fab expansion. This is peak-expectations mechanics, not fundamental weakness. The DRAM read-across is positive: 3Q ASPs rising well, semi equipment pricing leverage likely underappreciated heading into 2H. Watch $MU $LRCX $AMAT.

4/
$IBM -25% Tue on a surprise negative 2Q pre-announcement ($17.2B revenue vs. $17.9B consensus). The miss was execution: Transaction Processing down HSD, large deal slippage, Mainframe miss despite strong backlog. Distributed Infrastructure +37% YoY. The -26% reaction prices structural deterioration the data doesn't support. The more important read is the cyber commentary. IBM cited "rapidly-evolving, industry-wide cybersecurity concerns" as a demand headwind. Channel checks suggest this refers specifically to Anthropic's Mythos elevating the enterprise threat environment, more capable AI models expanding the attack surface and accelerating security budget urgency. IBM also reprioritized capex toward servers, storage, and memory including firewalls, a short-term positive for firewall demand, but some of that strength is pull-forward ahead of expected price increases. The $PANW thesis is reinforced, not threatened.

5/
$AAPL+5.8% on the week, one of the few large-cap bright spots. CAC approved Apple Intelligence in China on July 15, powered by Alibaba's Qwen, compressed from 54GB to under 4GB for on-device deployment. Apple held 18.1% of China smartphone market in Q2 2026 (up from 13.9% a year earlier, shipments +20%) even as the overall market contracted. The absence of Apple Intelligence in China had been a visible competitive gap for the past year. This closes it. Separately, Apple is reportedly qualifying $CXMT DRAM chips for future devices, supply chain diversification from Samsung/SK Hynix/MU. Combined with CXMT's 200x oversubscribed STAR Market IPO this week, Apple adoption would be the single most significant commercial validation of Chinese semiconductor capability to date.

#Semiconductors #AI #AIInfrastructure #Tech #Cyber

[Carl Quintanilla @carlquintanilla]

“.. This lofty level has only been reached two other times in history: two months before the peak of the Tech Bubble and four months ahead of the peak in 2007 that led the GFC.”

https://t.co/gJScXQs6Nh