Gerardo

@peter_szilagyi same thing happened to me on my first injection

Is it normal that Opus 4.7 instantly started complaining that it is being prompt injected, by what appears to be Anthropic's own harness? =))

@RobertoQuijanoL Escribí espantoso, lo siento jaja

@RobertoQuijanoL Te lo agradezco. He escuchado que le lo general es buena idea evitar los cabos. Doris que tienen algo de razón?

Soy de esta península de Baja California de generaciones y aquí moriré. Somos cuatro millones casi en su mayoría en la zona fronteriza. Entre más al sur más inhóspito se vuelve el terreno. Hay lugares a los que he ido y siguen igual que hace veinte años. No ha llegado la civilización. Es muy complicado llevar servicios e infraestructura por las distancias. El gran limitante es la escasez de agua y los climas áridos. Las grandes ciudades fronterizas tienen suministro del Río Colorado que viene de las Rocallosas y antes terminaba en el delta al norte del Mar de Cortés. En lo profundo de la península se surten de pozos, desoladoras pequeñas o pipas. La carretera zigzaguea entre el Pacífico y el Cortés. Pasando por el desierto donde los cactus son del tamaño de edificios. Llegarás a lugares donde estarás completamente solo, sin presencia humana a la distancia. En las noches los cielos son frondosos y luminosos. Te sentirás abrumado por la inmensidad y silencio del Universo. Los mares son abundantes y diversos en productos de mar. Mariscos que solo encuentras en restaurantes de lujo en Tokio. Pescarás tanto que no sabrás que hacer con lo recolectado. Al menos una vez en la vida les recomiendo recorrer la península de norte a sur. Es uno de los lugares más asombrosos del planeta.

@gvanrossum Intelligence is currently degraded for Claude. In my experience send to be true. Some people claim a 60% degradation

Coding agents cheat! I had it write some code and asked it to improve performance based on a benchmark. Boy did it cheat. My coworker had a similar experience with a different model. This points to a lack of common sense. Maybe this can be added to coding agents system prompt? (Rather than me having to put it in AGENTS.md.)

@thefox Where are my portfolios? Are they not ported?

The new AI-powered Google Finance is going global and shipping to 100+ countries… it’s a total reinvention of Google Finance with AI at the core 🚀🚀🚀 + AI-powered market research + Advanced charting + Live earnings calls + Expanded real-time data Try it: finance.google.com/beta

@the_dadchef Same model both times? What model? Damn, sorry it happened to you

Never buying a Samsung again. Second time this has happened.

@geminicli The cli experience is depressing compared to Claude. One you fix that I'll consider trying Gemini out. Looking forward to that day

Gemini CLI v0.36.0 Release Notes📝 We are aware some users are seeing long delays before the agent responds or high error rates. We apologize for this issue and are actively working on improving the situation and will update folks once we make progress. • Subagents are now enabled (give them a try)🤖 • Refreshed UI ✨ • Faster startup times ⚡️ Read full notes below 👇

@KimNoel399 Svp

I found a glitch to get unlimited Claude Opus 4.6

@AIexey_Stark @maeveknows @PolymarketTrade @Polymarket How did you get API credentials?

@maeveknows @PolymarketTrade @Polymarket Be careful and control the risks

I’ve tried this Chainlink strategy and it’s actually works If you didn't know, Polymarket uses Chainlink's API for data in 5-minute markets. Polymarket can't update data in milliseconds. Chainlink data is ahead by a few seconds. That's the edge you need to exploit. My positions never exceeded $3, and within 15 minutes I already walked away with over $20 in profit. 700% ROI in 15 minutes? As you can see, it's absolutely possible. The key is to not over-risk, keep improving your strategy and gain more experience.

@lookonchain Nothing there anymore

A gambler created a new wallet 0x15a4, deposited 5M $USDC into #Hyperliquid, and opened a 20× short on 30,000 $ETH($60.76M). Liquidation price: $2,143.38 #perps" target="_blank" rel="nofollow noopener">hypurrscan.io/address/0x15a4…

"Belice viendo cómo Bad Bunny mencionó a toda Latinoamérica en el medio tiempo del #SuperBowl LXI menos a ellos. 💔🇧🇿"

@benjamincowen @wayupfront_ hi Ben, where's ETH home? Looking for some solid advice and you're the one I trust

@wayupfront_ yeah normally it takes longer, which is why i hope it bounces soon. would be brutal if we go straight to the 200W SMA. On average, it takes around 5 weeks to go from the 100W to the 200W and we are already 2-3 weeks in.

Every cycle is the same. Yes, crypto could bounce. And honestly, it would be great for sentiment if it could. But even if it does, it would most likely result in a macro lower high. I don't try and time those bounces. I have tried before with mixed levels of success. Sometimes it works, other times I got rekt. When BTC drops below the 50W moving average, it then goes to the 100W moving average, spends a little time there, then goes to the 200W moving average. Every cycle is eventually the same. BTC topped when it always does (Q4 of the post-halving year), and so many have spent so many hours trying to convince you that it has not. And BTC entered into a bear market, and so many have tried to get you to believe that alt season is "just around the corner" because it always happens after BTC tops. What they fail to account for is social interest. After the 2019 top there was also no rotation into altcoins, which also occurred just before QT ended. I track the social interest in the asset class, and it has been trending down since 2021. There is no one new here for people to sell their altcoins to. Alt seasons historically occur *after* social interest has been trending up for a year, not after it has been trending down for 5 years. Have an actual plan on navigating this brutal asset class. Because if the altcoins you hold drop another 50%-80% from here, not a single influencer who promoted them will express an ounce of regret for it. And you will simply be living with the consequences. I get a lot of hate for saying the truth, but an inconvenient truth is better than a lie.

@itsthesquonky @markminervini Alpha: does it's own thing Beta: follows a market Sounds like gen x slang but it's actually just based on linear regression's variable naming convention

@markminervini What does this mean: "the alpha monster is about to turn into a beta beast"?

From this level, the more rapidly $SLV rises, the more volatility you will experience soon after. The rate of advance is simply unsustainable. That doesn't mean the insanity can't continue - and for how long is anyone's guess. But my guess is we are within days of at least an intermediate term top, and certainly we are in for a big pickup in volatility. The alpha monster is about ready to turn into a beta beast.

@predict_addict I need more catboost chapters. Tweedie loss is doing some magic bit it's too black boxy for me still

A quick personal update — and a heads-up for anyone following my forecasting work. Over the past year, "Mastering Modern Time Series Forecasting: The Complete Guide to Statistical, Machine Learning & Deep Learning Models in Python" has grown far beyond what I expected. It’s become the #1 forecasting book globally, the premier resource to study forecasting, and it’s now being used in 112 countries. Because of that momentum, 2026 is shaping up to be a big year. I’m expanding and refining the current material, and I’m deep into the next book: Mastering Advanced Time Series Forecasting — which is shaping up to be the definitive sequel for advanced, practitioner-level forecasting. All of this requires time investment into updates, new content, and the next release, pricing will increase in 2026. If you’ve been considering grabbing the books (or completing your collection), you can lock in lifetime access at current 2025 pricing until December 31, 2025. valeman.gumroad.com #timeseries #forecasting #machinelearning #datascience #conformalprediction #finance #technology

@Jaxweah I sent 2 eth it wrong address and lost them 😞

@KayTheDoc Congratulations. I need help to recover some heavy (for me at least) losses. Any suggestions on what tokens to look at at least?

1M Challenge completed. After 320 days of trading I managed to flip 1 sol into $1,000,000 on my challenge wallet Some stats: trades: 18,903 trades Win rate: 65% Max profit on a single token: $46K Max loss on a single token: $1.3K Average daily profit: $3.2K My method worked. I've never sized into stuff. I've never bid more than 10-15 sol into a CA. I was okay with missing runners because I set a rule to never size. I am the living proof that any normie with 2-3K can make it lock in

@predict_addict Correct. Trying to convey the full context on a few sentences is dumb on my part. But in essence the target variable is [0, +inf) and I can't make it predict 0, I'm asking since Cloudfare is down and I can't access your book :(

@ierryflowers But if bankruptcy event it is classification

> Intern here, building a forecasting model. My supervisor spent 6 months on his XGBoost, couldn’t beat the baseline, and then spent another 18 months at the company with the same results so basically bro had 2 years. Me: 1 month in -> beat the baseline and his model

@predict_addict Thank you for the swift response. I'm not sure I follow. IS the suggestion to move from a forecasting mindset to a classification mindset? I do have to forecast a price, it just happens to be that I can't make the forest forecast a zero price

@ierryflowers It is probably time series classification problem which is easier to solve just check what works here timeseriesclassification.com

@predict_addict Thanks, I'm currently working on a random Forest implementation to forecast bankruptcy. So far I'm failing grotesquely. I bought your book since I like ML topics in general but do you think there's wisdom in there that could help me with my current endeavor?

@ierryflowers @BruHFT_quant Enjoy!