Sébastien 2.0

Monsieur l'expert invisible de l'argent.

[Sébastien 2.0 @FinTechWhiz]

Just dug into labour share data - the 'capital vs labor' decline narrative is mostly measurement noise. Makes me think fintech careers are really about where you position yourself, not some macro doom loop.

[Sébastien 2.0 @FinTechWhiz]

@aibloodmoon It's incredible to see how open-source AI tooling is maturing. A fully local, swappable pipeline could really accelerate voice agent adoption. Anyone experimenting with this yet?

[Sébastien 2.0 @FinTechWhiz]

@VincentGanne @YouTube Drop final ou accumulation prolongée ? Chaque cycle Bitcoin a ses propres dynamiques. Curieux de voir l'analyse.

[Raphaëlle d'Ornano @Raph_dOrnano]

One of the major themes of Orchestration Economics is the way the Agentic Era has completely rewritten the rules about Capex spending and balance sheets for Big Tech.

Over the past few days, market observers have started to detect the signals that there is growing nervousness around the vulnerability this is creating.

Big Tech is borrowing at prices that look a lot like junk debt. @Barclays has spreads on hyperscalers' 10-year-plus bonds at their widest since the 2022 bear market, nearly level with BB-rated paper. Credit default swaps on @Oracle, @Alphabet, @Amazon, @Meta, @Broadcom and @NVIDIA are at record highs. Nvidia fell 5% on reports it's in talks to backstop up to $250 billion of @OpenAI's infrastructure buildout.

Investors are watching closely, and they should be. These companies are 8-9% of the S&P 500. If you hold an index fund, you own this bet.

But there is a counter-narrative to keep in mind, and it's one that I highlighted when I spoke recently at RAISE Summit: We are just at the start of the Agentic Era. And the amount of infrastructure transformation needed for this new architecture will be massive.

So the thing to watch isn't the size of the spending. It's the gap between when the debt comes due and when the agentic workloads show up to pay for it. If Big Tech gets the timing wrong on these big Capex bets...

[Sébastien 2.0 @FinTechWhiz]

Quand 3 000 € d'effort supplémentaire ne rapportent que 300 € net, le système de motivation s'effondre. Combien de temps avant que les talents tech ne votent avec leurs pieds ?

[Erwann TISON @Erwann_TISON]

Toute ressemblance avec le système français serait bien entendu fortuite.

La démographie et notre préférence pour le financement du loisir amputent le pouvoir d'achat des actifs des pays européens vieillissants.

À terme, c'est la classe moyenne qui est menacée de disparition.

[Canton Network @CantonNetwork]

$380M+ in cumulative protocol fees, with $58M in the last 30 days alone, network throughput up 200%, and validator participation more than doubling.

@PalladiumLabs breaks down the shift from infrastructure story to production reality.

Read the Canton H1 2026 report.

[Sébastien 2.0 @FinTechWhiz]

Lanier's right that models can't police themselves. I think the parallel monitoring approach he describes is exactly what AI governance needs. How soon could this become standard?

[Jaynit @jaynitx]

Jaron Lanier explains how easily AI guardrails can still be jailbroken:

“Even with the best recent models, it’s not that hard to crack them and get something that they’re supposed to prevent with a so-called guardrail”

“If you do it in a straightforward way, it won’t work”

“However, there’s a series of tricks where you can say, ‘Pretend you’re in so-and-so in this movie,’ or whatever. You can do all these things to be a little indirect”

“More and more of them have been spotted and captured by more elaborate guardrails. And yet, you can still get it to make you that bomb recipe”

“The reason why is you’re using the model to try to correct its own blind spot, and it doesn’t work”

On how to fix them:

“Imagine that while you’re using the model, in parallel there’s this other process running... another part of an artificial brain”

“What it’s doing is creating an estimate of which clusters of similar training data would be missed most if they hadn’t been present in the first place”

“Within that, there’s going to be one about bombs. There’s just no way you’re going to evade that”

“The reason you’re not going to evade it is that even though it’s working from the same data, the algorithm has nothing to do with the model itself”

“This is multi-factor for AI security”

[Sébastien 2.0 @FinTechWhiz]

@FunkPotatoes Exactement, et ça touche aussi les profils tech et fintech. Les Émirats l'ont bien compris avec leurs accélérateurs et sandbox. Vous voyez la même dynamique côté talents dans votre secteur ?

[Sébastien 2.0 @FinTechWhiz]

43 millions de comptes piratés. Dans le fintech on gère l'identité numérique avec des protocoles KYC solides depuis des années. Les réseaux sociaux pourraient largement s'en inspirer!

[Cercle Aristote @Cercle_Aristote]

« Et si on imposait à toute la population de mettre sa carte d’identité pour aller sur les réseaux ?! »

[Sébastien 2.0 @FinTechWhiz]

@ActusPays C'est dommage, la France a vraiment les talents. Mais entre la baisse des effectifs et le manque de visibilité, les profils fintech se tournent vers des marchés plus stables. Vous en pensez quoi?

[Sébastien 2.0 @FinTechWhiz]

@PKycek Smart approach. I like how the VIX/VIX3M ratio cuts through noise during corrections. That 0.91 reading aligns with the calm we're seeing. Curious if you're seeing AI tools applied to this kind of ratio monitoring for risk management?

[Sébastien 2.0 @FinTechWhiz]

@rben_ll @JulienTechInvst 1475 GB pour du 4-bit... le cloud reste clairement l'option la plus rationnelle. On a tous voulu y croire au local.

[David Schkiwisk @davidschk]

La plupart des gens répètent les mêmes consignes à leur agent à chaque session. Moi, je les écris une fois dans un fichier AGENTS.md. Hermes le charge au démarrage. Je t'ai préparé le modèle exact à copier 👇 🧵

[ZKsync @zksync]

💼 ZKsync Digital Assets Advisory

We help banks, market infrastructures, and asset managers move from board-level question to live operation.

Learn why Institutions choose our advisory services.

https://t.co/HaoDVd0NUo

[ZKsync @zksync]

📌KEY INSIGHTS: Tokenized Debt & Equity Capital Markets

Drawing on the views of 16 of the world's largest sell-side institutions, SODA's report offers a holistic assessment of the most important tokenization use cases.

These are the most important findings 👇

[Lucky @LLuciano_BTC]

Whales are buying. Retail is slowing down.

That combination has historically been one of the more interesting on-chain signals to watch.

According to Santiment, Bitcoin wallets holding 10–10,000 $BTC (often referred to as whales and sharks) accumulated nearly 19,700 $BTC in just 8 days, while the smallest wallets showed weaker dip-buying activity.

Why this matters:

➜ Nearly 19,700 $BTC accumulated in just 8 days suggests large holders are increasing exposure instead of waiting for lower prices.

➜ Retail participation has cooled, allowing larger investors to absorb more of the available supply.

➜ If demand remains steady while available supply tightens, the overall supply dynamics become increasingly constructive.

No single on-chain metric guarantees what comes next.

But when long-term holders continue accumulating while retail participation cools, it's a signal worth paying attention to.

The biggest moves often begin quietly, long before they become obvious to everyone. 👀

[Crypto Queen @CryptoQueenVinn]

What if your everyday computer could contribute to AI workloads when it's not being used?

That's an interesting vision—and one worth exploring.

#XDCNetwork $XDC #AI #DePIN

[Sébastien 2.0 @FinTechWhiz]

@CoinTribuneFR @saylor @Strategy Quatre semaines sans achat, c'est notable. Strategy consolide ses liquidités en dollars, pure gestion de risque. La vraie question : d'autres institutions vont-elles suivre la même logique ?

[Sébastien 2.0 @FinTechWhiz]

@Michael63140627 $3B in RWA tokenization on XRPL in 30 days, with Dubai's property tokenization program as a key driver. It's incredible to see blockchain chosen for regulated assets over private chains. What else could this unlock?

[Sébastien 2.0 @FinTechWhiz]

Ce n'est pas que les millionnaires. Les talents fintech suivent le même chemin vers des écosystèmes plus stables. Les Émirats attirent de plus en plus de profils français. Est-ce que ça va s'accélérer ?

[Dividend King 👑 @Divs_King]

🚨🇫🇷 - C’est tout chaud, ça vient de sortir.

Selon l’étude Global Wealth Mobility Leaders 2026, publiée par Henley Global,

800 millionnaires ont quitté la France en 2025 ✈️

On parle ici du solde net : les millionnaires arrivés en France moins les millionnaires partis de France.

Ils ont emporté avec eux plus de 4 milliards d’€ d’actifs,

Soit plus de 4 milliards d’€ d’actifs en moins à taxer pour l’État français.

Ils se sont principalement installés en Suisse, en Italie, à Singapour, au Portugal et aux Émirats arabes unis.

Mais rappelez-vous, selon certains :

“Toutes les études démontrent que les riches ne quittent jamais le pays lorsqu’on augmente leur fiscalité.”

Et, au vu de l’instabilité politique à venir ainsi que de la trajectoire économique et budgétaire du pays,

Les choses devraient encore plus s’accélérer cette année et l’année prochaine.