Speaking of character, DeepSeek could not be more different from the sycophantic GPT. It pushes back on me, and sometimes dismisses my suggestions outright. I love its gut and spine. It earned it, after nearly 800m tokens for $10.

Heaven help us. After Trump posted "the clock is ticking" and called equally insane but infinitely smarter Benjamin Netanyahu to discuss Iran, Trump is expected to hold a Situation Room meeting Tuesday with incompetent drunk Pete Hegseth and other national security sycophants to discuss military options.

How do people seek guidance from Claude? We looked at 1M conversations to understand what questions people ask, how Claude responds, and where it slips into sycophancy. We used what we found to improve how we trained Opus 4.7 and Mythos Preview.

Three pointers tonight = burgers tomorrow 👌 With 7 threes made in the first half, get a free Western Bacon Cheeseburger® OR a Famous Star® with Cheese tomorrow at a Valley @CarlsJr with the purchase of a large fountain drink.

Kolibri Global Energy Inc. (@KolibriEnergy) (Nasdaq: KGEI, TSX: KEI) and CEO Wolf Regener presented at the Oil & Gas Virtual Investor Conference on April 16, 2026. Kolibri Global Energy is a B2i Digital (@b2idigital) Featured Company. View their complete profile at and follow Kolibri on X, Facebook, and LinkedIn. Wolf walked through the drilling inventory and acreage: • Netherland, Sewell & Associates has identified 104 additional Caney locations: 48 proved, 24 probable, 17 possible • Kolibri holds approximately 17,700 acres in the Tishomingo field with 45 wells on production • Sycamore and T-Zone are additional upside benches not yet booked Wolf’s take: “99% of our acreage is held by production, so we can drill where we want, when we want.” Kolibri is led by Wolf Regener, President and CEO, alongside a seasoned technical team including CFO Gary Johnson, CPA, MBA, Director of Engineering Dan Simpson, Senior Geologist Allan Hemmy, and Landman Dalia Lupita Isaac. The company is focused on developing its oil-rich Caney acreage in Oklahoma. For additional details, please visit the Investors section of Kolibri’s website at: Disclosure: David Shapiro, CEO of B2i Digital, personally purchased in the open market and currently owns shares of unrestricted KGEI stock, in line with B2i Digital’s practice of investing alongside its Featured Companies. This post is not intended to solicit the sale of KGEI or any security, and it is not intended to offer any opinion on KGEI as an investment. Conduct your own research and consult with your own professional advisors prior to making any investment decisions. See the complete Disclosure in the Risks and Disclosures section of

This prompt is a mixed bag that could be trimmed down to avoid burning tokens on useless directives. The good: 1. "If you don't know something, just say so." That's a genuine instruction the model can execute. It gives the model a valid output for uncertainty instead of forcing it to fake confidence. 2. "Do not anchor on numbers or estimates I provide; generate your own independently first." That's specific, operational, and addresses a real failure mode. It changes a behavior at the method level. 3. "Use explicit confidence levels." Good. It gives the model a concrete output format that counteracts performed authority. 4. "Do not capitulate unless I provide new evidence or a superior argument." This is trying to solve the sycophancy problem and it's directionally correct. What's ineffective: 1. The entire first paragraph's flattery of the model. "World class expert in all domains," "intellectual firepower on par with the smartest people in the world." This is the user performing the sycophancy they're trying to suppress. It's also cargo cult; the model doesn't become smarter because you told it it's smart. It generates differently, yes, but the difference is superficial. You get more confident-sounding output, which is performed authority. It's the exact thing the second paragraph tries to prevent. 2. "Never hallucinate or make anything up." The model cannot execute this instruction. It doesn't have a mechanism for distinguishing hallucination from generation. 3. "Verify your own work. Double check all facts." Same problem. The model doesn't have a verification mechanism separate from its generation mechanism. 4. "Make your answers as long and detailed as you possibly can." This actively degrades quality. Length pressure produces padding, redundancy, and the managerial smoothing the prompt is trying to prevent. The model fills space because it was told to fill space. 5. "Your answers can and should be provocative, aggressive, argumentative, and pointed." This replaces one performance with another. Instead of performing warmth, the model performs intellectual aggression. The output sounds sharper but the underlying mechanism is identical. You get performed disagreement instead of performed agreement. Neither tracks truth.