flessner · HackerTrans

I gave it a test spin. Half an hour and the 5 hour usage cap was hit in Claude Code. Not what I would expect on the Max 20x usage plan. I am sure it is great, but at this rate I would rather finish what I am doing with Claude Opus instead of structuring my usage around the 5 hour windows.

flessner

4 個月前discuss

This is similar to building a React SPA and complaining that Google can't index it.

LLMs will use your website anyway. You're just choosing whether to pay the cost in structured endpoints upfront or hand that cost to browser emulation and lose control of how you're represented.

flessner

5 個月前discuss

No, commenters here simply take into account that predictions markets have historically classified themselves as futures markets and not as gambling.

Allowing information asymmetry, like insider trading, undermines the regulatory argument that keeps these markets legal.

flessner

5 個月前discuss

I am guessing there will be an OpenClaw "competitor" targeting Enterprise within the next 1-2 months. If OpenAI, Anthropic or Gemini are fast and smart about it they could grab some serious ground.

OpenClaw showed what an "AI Personal Assistant" should be capable of. Now it's time to get it in a form-factor businesses can safely use.

flessner

7 個月前discuss

If this is about military capability, why ban all foreign manufacturers, including proven innovators like Helsing and Baykar? Instead of blanket bans, targeted contracts could leverage Ukraine tested designs while building domestic capacity.

Innovation happens under competitive pressure. The US just created a domestic vacuum.

flessner

7 個月前discuss

That is exactly what you never want to do under protectionist policies. Domestic producers are shielded from Chinese competitors. This means they are under less pressure to reduce prices and innovate.

I wouldn't read too much into the national security justification. It's a political argument to an economic policy.

flessner

9 個月前discuss

I started learning the guitar years ago, but lost motivation once I got into university. Maybe I'll give it another shot and use this as a refresher on the theory!

Anyway, one small nitpick on the website: When on German language the word "FUNKTIONSHIGHLIGHTS" overflows on mobile. I would replace it with "WICHTIGSTE FUNKTIONEN" as that is two words.

Good luck, the website and App look nice!

flessner

10 個月前discuss

Out of curiosity, what made you pivot from a web/translation editor to recovering stripe payments? Was it primarily because of the teams prior experience in payment recovery or something else?

flessner

10 個月前discuss

As much as I like SEPA it is primarily for bank transfers.

The way that payments work through SEPA is that the merchant pulls the money from your account. Legally they require a "mandate" - this can be as little as a handwritten signature on a document.

Security is essentially provided by easy reversal and strong penalties for abuse.

flessner

10 個月前discuss

Vibe coding by itself isn't a problem.

The problem is vibe coding AND negligence. Good software practices like testing, code review, documentation are bound to catch the LLM-isms.

No offense on the author, the project specifically calls out that it's a "young" project in the footer, so I personally wouldn't expect it to be quite up to spec yet.

flessner

11 個月前discuss

When you look at the market with a zero-sum perspective it becomes apparent that both active and passive investors earn average market returns - collectively they by definition are the market.

However, active investors have higher trading fees/management costs, so they are bound to perform at least slightly worse on average. It's just mathematics.

flessner

11 個月前discuss

That's a common pattern in trading strategies with negative skews or tail risks. Even large hedge funds, like LTCM, can fall into this same pitfall.

For anyone interested, I can recommend the book "Systematic Trading" by Robert Carver. You don't have to be into algorithmic trading, the sections on risk management and positive vs negative skews are already worth the read.

flessner

去年discuss

Interesting question, here are my few cents.

Exports are more "expensive" for countries with reserve currency status. This is a problem for countries that export many primary and "low-tech" secondary sector products. Countries that export many "high-tech" secondary sector products can usually still thrive.

This leaves us with the usual suspects: US, China, Germany (-> EU) and Japan.

flessner

去年discuss

I get it that "click here" is not descriptive, but so is simply linking "Amaya". What is it? A person? A fruit?

People don't read websites linearly, in the best case they skim read all the buttons and links. I personally would include the verb as it gives important context and is a clearer CTA for the "skimmers".

Amaya is W3C's... "Download Amaya"!

flessner

去年discuss

From an economic perspective it requires LLMs and humans to have comparable outputs. That's not possible in all domains - at least in the near future.

flessner

去年discuss

> Already we live with incredible digital intelligence, and after some initial shock, most of us are pretty used to it. Very quickly we go from being amazed that AI can generate a beautifully-written paragraph to wondering when it can generate a beautifully-written novel;

It was probably around 7 years ago when I first got interested in machine learning. Back then I followed a crude YouTube tutorial which consisted of downloading a Reddit comment dump and training an ML model on it to predict the next character for a given input. It was magical.

I always see LLMs as an evolution of that. Instead of the next character, it's now the next token. Instead of GBs of Reddit comments, it's now TBs of "everything". Instead of millions of parameters, it's now billions of parameters.

Over the years, the magic was never lost on me. However, I can never see LLMs as more than a "token prediction machine". Maybe throwing more compute and data at it will at some point make it so great that it's worthy to be called "AGI" anyway? I don't know.

Well anyway, thanks for the nostalgia trip on my birthday! I don't entirely share the same optimism - but I guess optimism is a necessary trait for a CEO, isn't it?

flessner

去年discuss

I recently got a new laptop and had to setup my IDE again.

After a couple hours of coding something felt "weird" - turns out I forgot to login to GitHub Copilot and I was working without it the entire time. I felt a lot more proactive and confident as I wasn't waiting on the autocomplete.

Also, Cursor was exceptional at interrupting any kind of "flow" - who even wants their next cursor position predicted?

I'll probably keep Copilot disabled for now and stick to the agent-style tools like aider for boilerplate or redundant tasks.

flessner

去年discuss

I have been interested in algorithmic trading for quite a while now. Everything you've said resonates with me as I ran into similar issues. I hope you don't mind that I add my own couple cents here.

(a-c) LLMs are especially difficult to use due to their knowledge cutoffs and "unpredictability". A self-trained "old-school" machine learning model can go a long way though.

(d) With Crypto the volatility is great for trading, but liquidity can quickly become a problem (even at $1000 non-leveraged positions). For me, the ultimate goal is to find a strategy that is profitable in all market conditions. I personally value consistency and reliability more than absolute profit.

(e) There's some chatting about risk management, but absolutely no discussion on profitable strategies. Resources are incredibly scarce - Systematic Trading by Robert Caver is the only book that was actually useful.

flessner

去年discuss

Absolute beginners usually don't even know how to ask proper questions. LLMs can help you in that regard... they'll answer your questions and provide you an answer no matter how trivial.

However, over reliance on it - like with all technologies - doesn't end well.

flessner

去年discuss

On iOS there's also Beorg [1], which I used briefly.

The problem that org-mode and Markdown face is that they are fairly minimal. It's common to use plugins, that aren't supported everywhere. This kills portability, which is a core "selling point".

I have seen it happen it Obsidian and Logseq - but even GitHub has a slightly altered Markdown spec.

[1] https://www.beorgapp.com