#prediction

• Thu, Oct 2, 2025. Forecasts give comfort. They may not be good but they feel safer than instinct. The Knowledge Project: Barry Diller #future #prediction

• Thu, Oct 2, 2025. Create a web app that polls a Slido survey and runs topic modelling to visually show the clusters of questions live, e.g. like pol.is. #prediction #hard

• Mon, Sep 22, 2025. IDEA AI settings that maximize persuasion #future #prediction

• Wed, Sep 10, 2025. LLMs are non deterministic because GPUs add floating point numbers concurrently and FP addition is non associative - order matters. Thinking Machines #gpu #llm-ops #prediction

• Tue, Sep 2, 2025. 2 Sep 2025 #best-practices #code-agents #future #prompt-engineering #prediction
  • Bias detection on declined applications tied to regulatory scrutiny (concrete, high-value, demoable on synthetic data).
    • Compute fairness metrics (selection rate, demographic parity diff, equalized odds) by protected_group; show SHAP top features per group; propose three policy levers to reduce disparity.
  • Scenario “adverse action” simulator (loss/collections outcomes vs policy levers).
    • Inputs: threshold, line increase/decrease, collections intensity; Outputs: approval rate, ECL, vintage loss, roll-rate curves; include confidence bands.
  • Self-learning Prompt → SQL log
    • Capture each natural-language question + generated SQL + exec time + rowcount → audit table; proves control & learning loop.

• Mon, Aug 25, 2025. IDEA Counterfactual notebook #future #prediction
  • Rewrite notes replacing "must" with "might" or "will" with "won't". See what reasoning is fragile & falls apart

• Fri, Aug 22, 2025. DSPy auto-optimizes prompts based on input-output pairs or evals. Typical improvements are ~10-20%. My opinion: avoid. It's a good idea, but has too much abstraction that hides the implementation. Worth learning from but not implementing. #automation #dev #gpu #markdown #optimization #prompt-engineering #prediction

• Mon, Aug 18, 2025. can make us overconfident. (At least, it makes me overconfident.) They create surprisingly good output, but only ~20% of the time. I cannot commit to a specific task based on that. Instead, it's better to rely on AI coding estimates for portfolios, e.g. promise to share something cool without mentioning what. Or do something cool first, then share. #ai-coding #code-agents #future #prompt-engineering #prediction

• Sun, Aug 3, 2025. Research indicates that we don't know in advance which prompts will help. Evals beat prompt engineering. Ethan Mollick #future #prompt-engineering #prediction

• Wed, Feb 26, 2025. When evaluating inputs, models tend to prefer the first response, prefer their own response, and prefer longer responses. ThursdAI #future #prediction

• Fri, Jan 17, 2025. Possibility Thinking is "the willingness to see possibilities everywhere instead of limitations". It's an approach / mindset that can make things that seem hard possible. With LLMs, this is becoming increasingly realistic to me in many areas. #future #llm-ops #prediction #think

• Thu, Dec 19, 2024. o1-preview diagnoses better than doctors. Harvard #future #prediction

• Thu, Oct 31, 2024. VibeCheck can predict a model based on its vibes 80% of the time. #future #prediction

• Sun, Oct 6, 2024. Jake Heller: #future #llm-ops #prediction
  • "One of the things we learned is (an LLM app) after it passes passes frankly even 100 tests, the odds that it will do, on any random distribution of user inputs, the next 100,000 100% accurately is very high."

• Tue, Aug 27, 2024. Cursor uses speculative edits and a variety of other techniques to speed up code editing. #ai-coding #future #markdown #prompt-engineering #prediction

• Tue, Jun 25, 2024. Rouge and Context Precision are metrics to evaluate LLM responses that serve as a starting point -- but not sufficient, usually #llm-ops #prediction

• Tue, Mar 19, 2024. Ways to use inversion #best-practices #dev #future #prompt-engineering #prediction
  • "Pre-mortem" is an analysis at the beginning of how a project failed. Then avoid that
  • "Red team" or "Black hat" are designated to contradict.

• Tue, Mar 19, 2024. Having a PoV IS a hypothesis. Always having a PoV allows us to detect anomalies and learn. #future #prediction

• Sun, Feb 4, 2024. As automation takes over these mainstream activities, people will take over the niches. Since expertise like knowledge is fractal, there will be many more segments of one in the future and it will be easier to automate clusters of similar abilities. Recommenders and brands will become even more important #automation #future #prediction

• Sun, Feb 4, 2024. Stephen Osserman's Observables have some nice notes. #markdown #prediction
  • Visualizing partial election results
  • D3 Force Dilemmas: Data Distortion

• Tue, Jan 16, 2024. As humans have more conversations with LLMs, they will replace video watching and interactive gaming with conversation based role play. New game genres will evolve #future #llm-ops #learning #prediction

• Sun, Jan 14, 2024. Practical AI podcast: AI predictions for #future #prediction

• Sat, Nov 25, 2023. Quote from Jerry Liu: "GPT 4 is really good at complex reasoning". It's worth exploring what that means. #chatgpt #future #gpu #prediction