Christian Marquay

12.3K posts

Christian Marquay

@cmarquay

Bank data engineer, Web3 enthusiast. Decoding finance with bytes and blocks. Let's shape the future of data together! 🚀 #DataEngineer #Web3InFinance

Katılım Kasım 2013

252 Takip Edilen614 Takipçiler

Christian Marquay@cmarquay·1 Nis

Google warns quantum computing could break crypto sooner than expected by reducing the resources needed to crack today’s encryption. Not a current threat - but a real future one. The takeaway: crypto isn’t broken, but the clock is ticking to go quantum-safe.

English

Christian Marquay retweetledi

Jerry Liu@jerryjliu0·31 Mar

i'm so cooked

rahat@Rahatcodes

Claude Code has a regex that detects "wtf", "ffs", "piece of shit", "fuck you", "this sucks" etc. It doesn't change behavior...it just silently logs is_negative: true to analytics. Anthropic is tracking how often you rage at your AI Do with this information what you will

English

37.3K

Christian Marquay retweetledi

nic carter@nic_carter·31 Mar

Many are wondering "what Google saw" that caused them to revise their post-quantum cryptography transition deadline to 2029 last week. It was this: research.google/blog/safeguard…

English

962

7.4K

Christian Marquay retweetledi

Garry Kasparov@Kasparov63·30 Mar

Novel environments, no precedents or plagiarism possible. Humans 100%, AI <1%.

Guri Singh@heygurisingh

Humans: 100% Gemini 3.1 Pro: 0.37% GPT 5.4: 0.26% Opus 4.6: 0.25% Grok-4.20: 0.00% François Chollet just released ARC-AGI-3 -- the hardest AI test ever created. 135 novel game environments. No instructions. No rules. No goals given. Figure it out or fail. Untrained humans solved every single one. Every frontier AI model scored below 1%. Each environment was handcrafted by game designers. The AI gets dropped in and has to explore, discover what winning looks like, and adapt in real time. The scoring punishes brute force. If a human needs 10 actions and the AI needs 100, the AI doesn't get 10%. It gets 1%. You can't throw more compute at this. For context: ARC-AGI-1 is basically solved. Gemini scores 98% on it. ARC-AGI-2 went from 3% to 77% in under a year. Labs spent millions training on earlier versions. ARC-AGI-3 resets the entire scoreboard to near zero. The benchmark launched live at Y Combinator with a fireside between Chollet and Sam Altman. $2M in prizes on Kaggle. All winning solutions must be open-sourced. Scaling alone will not close this gap. We are nowhere near AGI. (Link in the comments)

Català

138

307

2.4K

570.7K

Christian Marquay@cmarquay·30 Mar

AI isn’t your brain. It’s your exoskeleton. It doesn’t replace what you know. It amplifies it. Less time searching. More time building. Less hesitation. More iteration. Same you. Just faster, sharper, and harder to stop.

English

Christian Marquay retweetledi

Alex Freberg@Alex_TheAnalyst·26 Mar

Technical skills are 20% of the job... but you 100% have to have them to get most jobs in data. Here are the other 80% of skills that actually matter once you get the job: - Turning vague questions into clear problems you can actually solve - Knowing when “good enough” is better than perfect - Communicating insights in a way non-technical people understand - Asking better questions than anyone else in the room - Understanding the business behind the data - Managing stakeholders (and their expectations) - Prioritizing what actually matters vs. what’s just noise - Being reliable and hitting deadlines consistently - Documenting your work so others can use it - Not overcomplicating things just to look smart You still need the hard-skills (and they're important for getting a job), but in the real world they aren't 100% of what you need.

English

182

9.7K

Christian Marquay retweetledi

The White House@WhiteHouse·25 Mar

The goal is not just to reach the Moon, but to stay. 🌘

NASA Administrator Jared Isaacman@NASAAdmin

To return Americans to the Moon, NASA is shifting to an iterative, execution-focused approach – just as we did during Apollo. We are standardizing rocket architecture, embedding NASA expertise across industry, and increasing launch cadence to support sustained lunar operations. We are sending a demand signal for crewed missions beyond Artemis V, with at least two providers capable of bringing astronauts to the surface every 6 months. The goal is not just to reach the Moon, but to stay. America will never give up the Moon again.

English

4.6K

4.7K

29.1K

4.8M

Christian Marquay retweetledi

Andrej Karpathy@karpathy·28 Mar

- Drafted a blog post - Used an LLM to meticulously improve the argument over 4 hours. - Wow, feeling great, it’s so convincing! - Fun idea let’s ask it to argue the opposite. - LLM demolishes the entire argument and convinces me that the opposite is in fact true. - lol The LLMs may elicit an opinion when asked but are extremely competent in arguing almost any direction. This is actually super useful as a tool for forming your own opinions, just make sure to ask different directions and be careful with the sycophancy.

English

1.7K

2.4K

31.2K

3.4M

Christian Marquay@cmarquay·25 Mar

Quantum computing won’t kill banks or Web3 - it will expose who’s ready. Banks can upgrade quietly. Web3 must coordinate loudly. The real risk? Data harvested today, decrypted tomorrow. This isn’t hype. It’s a race against time.

English

Christian Marquay retweetledi

Bark@barkmeta·23 Mar

Let me explain what just happened 👇 5 minutes before the President announced a halt to attacks on Iran… someone placed a $1.5 BILLION bet on stocks going up and dumped $192 million in oil. 5 minutes… These trades were 4 to 6 times larger than anything else in the entire market. Whoever did this wasn’t guessing. You don’t risk $1.5 billion on a hunch. There was zero public indication this announcement was coming. No leaks. No press. Nothing. The only people who knew were in the room when the decision was made. Someone in that room picked up a phone. And within minutes they made more money than most Americans will earn in a thousand lifetimes. In a single trade. On a war that cost you $4+ a gallon gas and $16 billion in tax dollars. American citizens funded this war. Politicians are profiting from it. This is not the first time. Every major announcement from this administration has had massive suspicious trades right before it dropped. Tariff reversals. Policy shifts. War decisions. This is the most blatant insider trading operation in the history of American politics. It’s not even close. And it’s happening over and over in broad daylight. You would go to federal prison for trading on a tip from your cousin. These people are front running war decisions with billion dollar bets and nobody will ever ask a single question. Nobody will be investigated. Nobody will be charged. By tomorrow this will be buried under the next satisfying headline. Just like last time. And the time before that. The game is rigged. And they’re not even trying to hide it anymore…

unusual_whales@unusual_whales

BREAKING: Just five minutes before Trump's announcement to halt the attacks on Iran, massive trades reportedly hit the market. In one move, $1.5 billion in S&P 500 (ES) futures was bought while $192 million in oil (CL) futures was sold. These orders were 4–6x larger than anything else at the time. The trader seemingly made huge gains. Unusual.

English

59.1K

158K

12.2M

Christian Marquay@cmarquay·23 Mar

@PowerHasheur Pourquoi se fatiguer à prendre les charts depuis vendredi et à parler de gap CME quand on peut faire des spurious correlations ?

Français

Hasheur@PowerHasheur·23 Mar

BRICS News@BRICSinfo

BREAKING: 🇺🇸🇮🇷 President Trump says the US and Iran have had a "very good and productive conversations" over the last two days for a total resolution on the war.

ZXX

384

102.6K

Christian Marquay retweetledi

trade.xyz@tradexyz·18 Mar

S&P Dow Jones Indices and trade[XYZ] have joined forces to launch the first official S&P 500 perpetual contract, available exclusively on Hyperliquid. For 69 years, the S&P 500 has been a defining reference point for global finance. Until now, access to that benchmark has been shaped by market hours, intermediaries, and geography. Today, that changes. The S&P 500 perp is now available 24/7/365, anchored by the official index data required for deep liquidity and institutional confidence at scale. SPDJI helped define modern indexing. They are stewards of an iconic benchmark, the standard against which portfolios across the globe are measured. We are honored to bring that legacy on-chain. Trade[XYZ] is bringing the world's most iconic assets towards a future of global, continuous markets — a future powered by Hyperliquid.

English

930

1.9K

3.3M

Christian Marquay retweetledi

Netflix@netflix·17 Mar

Hudson Williams, Jennifer Grey, Terry Chen, Elizabeth Adams, Hannah Galway, and William Mapother join the cast of THE ALTRUISTS. The series tells the story of Sam Bankman-Fried (Anthony Boyle) and Caroline Ellison (Julia Garner), two hyper-smart, ambitious young idealists who tried to remake the global financial system in the blink of an eye...before they were accused of stealing $8 billion.

English

523

22.8K

2.2M

Christian Marquay retweetledi

vitalik.eth@VitalikButerin·10 Mar

I actually do the whole new year's resolutions thing, and it actually works. The key thing to understand is that humans are creatures of habit. Doing the same action you've already done regularly takes very little mental effort, whereas inserting a new one-time task takes much more. And so if you want yourself to do certain things more, you need to make it a habit. The year boundary is as good a place as any to evaluate the habits that you're chosen to impose on yourself, and see whether they are effective and sustainable, and adjust, add or remove any. My style is to make them measurable, trackable, and targeted to exactly the level of effort that I know will not make me want to abandon them, even during my months of busiest work, most intensive travel schedule or call schedule, etc. Examples I've done: * Walk an average of >= 6km/day each month * Run >= 50km each month * Write >= 1 blog post each month * Study some language for 30 min each week * Do >= 2 major cryptography programming projects each year At every year boundary, re-evaluate your old list, and decide on your new list. And yeah I have txt files for tracking this (sorry, not gonna use some corposlop app that makes me dependent on third-party servers) You actually want each one to be relatively trivial, so that you can stack multiple, and because the benefits of maximizing are less important than the risk that you will give up on the whole thing. This has worked well for me and I recommend it.

English

347

140

1.9K

161.3K

Christian Marquay@cmarquay·10 Mar

Vincent Courboulay (@vcourbou) warns that AI has two possible futures: “Like Dr Jekyll and Mr Hyde, AI can become responsible and sustainable… or destructive and polluting.” The real question isn’t how powerful AI becomes, but what we choose to use it for.

English

Christian Marquay@cmarquay·5 Mar

Vibecoding isn’t mysticism. It’s the ability to intentionally shape atmosphere through small cues: light, sound, pacing, texture, color. It’s emotional UX for real life. As AI, hybrid work, and experiences evolve, this skill becomes a quiet superpower.

English

Christian Marquay@cmarquay·3 Mar

Fast mode for speed. Thinking mode for depth. Smart teams use both: quick drafts, fast answers, instant exploration… then switch to deep reasoning for decisions that actually matter. Dual-mode AI isn’t a gimmick. It’s workflow strategy.

English

Christian Marquay retweetledi

NASA@NASA·28 Şub

The cosmos gifted us a rose 🌹 This rose is made of a pair of interacting galaxies that have been distorted by each other’s gravitational pulls. Learn more about these cosmic companions: science.nasa.gov/asset/hubble/a…

English

1.2K

7.6K

31.5K

1.7M

Christian Marquay retweetledi

Alex Freberg@Alex_TheAnalyst·18 Şub

I've been using Claude Cowork for the past few days with my real work and here's my honest review: 1. It's fast and thoughtful - Processes in-depth drafts what would take me hours - It breaks complex problems into steps and assess each step - Can create entire documents with good output 2. Context is difficult for longer durations - As I used it more it lost the context fairly quickly over time which I needed to continue to prompt - Not huge issue, but the longer the task it would often drift and do 90% what I initially wanted 3. It can give A LOT of output at one time! - This is good and bad. Lots of work done at one time, but also a lot to review! - And reviewing was very important for large documents it produced (got it right 90% of the time, some less if it didn't understand what I really wanted) - Downside was if it wasn't good, took a long time to run and redo 4. It deleted some of my files - I gave it access to some Folders I set up just for testing with some real files - It deleted my files and told me it didn't and was still there (not cool - luckily I had backups) 5. Connectors, Plugins, Skills are all great! - But they don't always work. In fact 25% of the time they just failed and Clause Cowork would have to find a workaround - It's a great ecosystem, but the ecosystem is not super reliable 6. It's good, but a lot of the online attention is Hype - Listen, it is good - absolutely, but it's not fire the entire Marketing, Coding, and HR teams good - Hype gets people trying it (like me) and I see many use cases for it in my personal uses, but it doesn't match all the hype see I like trying new things and staying up to date in AI and I think I'll continue to use Claude Cowork for a few use cases - definitely worth trying out!

English

6.4K

Christian Marquay@cmarquay·25 Şub

Wondering who will fight the ninja robots.

vitalik.eth@VitalikButerin

It will significantly increase my opinion of @Anthropic if they do not back down, and honorably eat the consequences. (For those who are not aware, so far they have been maintaining the two red lines of "no fully autonomous weapons" and "no mass surveillance of Americans". Actually a very conservative and limited posture, it's not even anti-military. IMO fully autonomous weapons and mass privacy violation are two things we all want less of, so in my ideal world anyone working on those things gets access to the same open-weights LLMs as everyone else, and exactly nothing on top of that. Of course we won't get anywhere close to that world, but if we get even 10% closer to that world that's good, and if we get 10% further that's bad) CC @DarioAmodei firefly.social/post/bsky/pv7f…

English

Keşfet

@PowerHasheur @elonmusk @BarackObama @taylorswift13 @cristiano @BillGates @NASA @nikifrancismediavine