<?xml version="1.0" encoding="utf-8" standalone="yes"?>
<rss version="2.0" xmlns:atom="http://www.w3.org/2005/Atom" xmlns:content="http://purl.org/rss/1.0/modules/content/">
  <channel>
    <title>whisper on S Anand</title>
    <link>https://www.s-anand.net/blog/tag/whisper/</link>
    <description>Recent content in whisper on S Anand</description>
    <generator>Hugo -- 0.156.0</generator>
    <language>en-us</language>
    <lastBuildDate>Sun, 24 May 2026 00:00:00 +0000</lastBuildDate>
    <atom:link href="https://www.s-anand.net/blog/tag/whisper/index.xml" rel="self" type="application/rss+xml" />
    <item>
      <title>Things I Learned - 24 May 2026</title>
      <link>https://www.s-anand.net/blog/things-i-learned-24-may-2026/</link>
      <pubDate>Sun, 24 May 2026 00:00:00 +0000</pubDate>
      <guid>https://www.s-anand.net/blog/things-i-learned-24-may-2026/</guid>
      <description>&lt;p&gt;This week, I learned:&lt;/p&gt;
&lt;ul&gt;
&lt;li&gt;BitWarden seems to be sneakily jacking up prices and going towards a PE sale. Might be time to shift out or self host. Sigh, I just migrated into it&amp;hellip; &lt;a href=&#34;https://blog.ppb1701.com/the-quiet-renovation-at-bitwarden&#34;&gt;Source&lt;/a&gt;&lt;/li&gt;
&lt;li&gt;Andrej Karpathy has joined Anthropic. Likely to use Claude to build better Claudes - automating AI research. Also, it probably isn&amp;rsquo;t a good time to build an AI education platform. &lt;a href=&#34;https://claude.ai/share/f9932e93-a632-4015-93f4-84359670f53c&#34;&gt;Claude&lt;/a&gt;&lt;/li&gt;
&lt;li&gt;The open-source Chinese models about 6 months behind frontier models. &lt;a href=&#34;https://qwen.ai/blog?id=qwen3.7&#34;&gt;Qwen 3.7-Max&lt;/a&gt; is &lt;a href=&#34;https://arena.ai/leaderboard/text&#34;&gt;on par&lt;/a&gt; with Claude 4.5 Opus (Nov 2025) and Gemini 3 Flash (Dec 2025).&lt;/li&gt;
&lt;li&gt;&lt;a href=&#34;https://blog.google/products-and-platforms/products/search/search-io-2026/&#34;&gt;Google basically became Gemini&lt;/a&gt;. Entirely! I&amp;rsquo;m not sure there&amp;rsquo;s a difference any more. Which means it will scrape websites and not send traffic through - just killing the search economy. But it&amp;rsquo;s far more useful. &lt;a href=&#34;https://claude.ai/share/9f3f6172-2965-40a0-8b67-053a0769e455&#34;&gt;Claude&lt;/a&gt;&lt;/li&gt;
&lt;li&gt;I wanted a list of sites I log into with my Google Account. &lt;a href=&#34;https://myaccount.google.com/connections&#34;&gt;Google&amp;rsquo;s Linked apps&lt;/a&gt; page does that. Unfortunately, I can&amp;rsquo;t find a way to use &lt;a href=&#34;https://takeout.google.com/&#34;&gt;Google Takeout&lt;/a&gt; to export that data. So I wrote a &lt;a href=&#34;https://github.com/sanand0/scripts/blob/deb4c1ecbc93e03511ca264ce14d2977d01b7d90/googleconnections.py&#34;&gt;scraper&lt;/a&gt; which can be &lt;a href=&#34;https://github.com/sanand0/scripts/blob/deb4c1ecbc93e03511ca264ce14d2977d01b7d90/prompts/googleconnections.md&#34;&gt;single-shot prompted&lt;/a&gt; these days.&lt;/li&gt;
&lt;li&gt;As long as you remember to exhale, your chances of recovery from being ejected into space is pretty good for the first 15-60 seconds. &lt;a href=&#34;https://gemini.google.com/share/4a15b461a2d3&#34;&gt;Gemini&lt;/a&gt;&lt;/li&gt;
&lt;li&gt;I don&amp;rsquo;t understand half the comments I read on LinkedIn. Earlier, I was able to separate good from bad. Now, I&amp;rsquo;m not sure if what I read is actually insight or idiocy. Is the AI use making their comments too smart or making my brain too dumb?&lt;/li&gt;
&lt;li&gt;&amp;ldquo;Pax Memoriae&amp;rdquo;: peace of memory. Putting past conflicts to rest. The best part of it was, I learnt the phrase by typing &amp;ldquo;Pax&amp;rdquo; into VS Code and wasn&amp;rsquo;t sure what to write next. Before I could search for it, GitHub Copilot completed it. I searched for what it meant, and it was &lt;em&gt;so apt&lt;/em&gt;!&lt;/li&gt;
&lt;li&gt;Children&amp;rsquo;s vision is worse than adults, but filter less and absorb ore irrelevant information than adults. This is useful for learning and surprise detection, but costly for focus, speed, and relevance. &lt;a href=&#34;https://chatgpt.com/share/6a0aae2f-9b64-83ec-8db5-00c2f3a465b0&#34;&gt;ChatGPT&lt;/a&gt;&lt;/li&gt;
&lt;li&gt;The word phobia comes from the Greek god of fear, Phobos, which is the name of one of Mars&amp;rsquo; moon. Deimos, the other moon, is the Greek god of dread/terror. They&amp;rsquo;re the children of Ares (Mars), the god of war. Nice planet.&lt;/li&gt;
&lt;li&gt;On WhatsApp, I can type &lt;code&gt;@Meta AI&lt;/code&gt; and then &lt;code&gt;/imagine&lt;/code&gt; to have it draw an image. The quality is OK - not great, not terrible.&lt;/li&gt;
&lt;li&gt;Surprising but GPT Realtime Whisper (&lt;a&gt; new model) isn&amp;rsquo;t as good as the older open-source Whisper models. Also, Gemini 3 Flash Preview is as good at transcription as Gemini 3.1 Pro Preview for up to medium-length text. &lt;a href=&#34;https://pythonicvarun.github.io/llm-audio-transcription-benchmark/&#34;&gt;LLM Audio Transcription benchmark&lt;/a&gt;&lt;/li&gt;
&lt;li&gt;Google Maps typically shows me a cycling time of 30 minutes when it take me 40 minutes and a walking time of 40 minutes when it take me 30 minutes. Either I walk much faster and cycle much lower than the typical person or Google Maps is not well calibrated to Singapore and India.&lt;/li&gt;
&lt;/ul&gt;
</description>
    </item>
    <item>
      <title>Things I Learned - 02 Mar 2025</title>
      <link>https://www.s-anand.net/blog/things-i-learned-02-mar-2025/</link>
      <pubDate>Sun, 02 Mar 2025 00:00:00 +0000</pubDate>
      <guid>https://www.s-anand.net/blog/things-i-learned-02-mar-2025/</guid>
      <description>&lt;p&gt;This week, I learned:&lt;/p&gt;
&lt;ul&gt;
&lt;li&gt;&lt;a href=&#34;https://www.proxmox.com/en/products/proxmox-virtual-environment/overview&#34;&gt;Proxmox Virtual Environment&lt;/a&gt; is an open-source alternative to VMWare, Hyper-V, Citrix XenServer, etc. (There&amp;rsquo;s nothing there that prompts me to explore it further.)&lt;/li&gt;
&lt;li&gt;With Podman on Windows (a Docker equivalent), many Docker-enabled tasks become easier. For example, running PostgreSQL is as easy as:
&lt;div class=&#34;highlight&#34;&gt;&lt;pre tabindex=&#34;0&#34; class=&#34;chroma&#34;&gt;&lt;code class=&#34;language-sh&#34; data-lang=&#34;sh&#34;&gt;&lt;span class=&#34;line&#34;&gt;&lt;span class=&#34;cl&#34;&gt;podman run -d --name postgres -e &lt;span class=&#34;nv&#34;&gt;POSTGRES_PASSWORD&lt;/span&gt;&lt;span class=&#34;o&#34;&gt;=&lt;/span&gt;postgres -p 5432:5432 postgres:latest
&lt;/span&gt;&lt;/span&gt;&lt;span class=&#34;line&#34;&gt;&lt;span class=&#34;cl&#34;&gt;podman &lt;span class=&#34;nb&#34;&gt;exec&lt;/span&gt; -it postgres psql -U postgres -c &lt;span class=&#34;s2&#34;&gt;&amp;#34;CREATE DATABASE mydb;&amp;#34;&lt;/span&gt;
&lt;/span&gt;&lt;/span&gt;&lt;/code&gt;&lt;/pre&gt;&lt;/div&gt;&lt;/li&gt;
&lt;li&gt;Bad deep research prompts are: vague/broad, under-specified or ambiguous. In short, the more you know what you want, the better. Iterate until then.&lt;/li&gt;
&lt;li&gt;What kind of reports do clients are research companies to produce? I was curious to see if Deep Research can replace these. Here are a bunch of ideas. &lt;a href=&#34;https://chatgpt.com/share/67bf7946-c80c-800c-8132-6f4018455a68&#34;&gt;ChatGPT&lt;/a&gt;
&lt;ol&gt;
&lt;li&gt;Strategy &amp;amp; Management Consulting Research (McKinsey &amp;amp; Company, Boston Consulting Group, Bain &amp;amp; Company, Strategy&amp;amp;, Accenture Strategy)
&lt;ul&gt;
&lt;li&gt;Produce a comprehensive strategic transformation report for a Fortune 500 consumer goods company. Analyze global market trends, competitor strategies, and actionable growth recommendations, including case studies and source citations.&lt;/li&gt;
&lt;li&gt;Generate an in‐depth study on corporate restructuring trends in emerging markets. Focus on successful turnaround strategies, CEO leadership factors, and strategic pivots, with a comparative analysis of key players.&lt;/li&gt;
&lt;li&gt;Create a report on M&amp;amp;A trends in the technology sector over the past five years. Detail deal drivers, integration best practices, and forecast future acquisition opportunities, citing relevant data.&lt;/li&gt;
&lt;/ul&gt;
&lt;/li&gt;
&lt;li&gt;IT &amp;amp; Technology Research Analysts (Gartner, Forrester Research, IDC, 451 Research, Ovum)
&lt;ul&gt;
&lt;li&gt;Produce a market assessment report on emerging cloud computing platforms. Include vendor evaluations, adoption forecasts, and key technology drivers with supporting data and charts.&lt;/li&gt;
&lt;li&gt;Generate an in‐depth cybersecurity trends report for enterprise IT. Analyze recent threat vectors, defense strategies, and best practices for risk mitigation, providing actionable recommendations.&lt;/li&gt;
&lt;li&gt;Create a comprehensive study on the impact of artificial intelligence in enterprise software. Include competitive benchmarking, technology adoption rates, and forecasted market changes.&lt;/li&gt;
&lt;/ul&gt;
&lt;/li&gt;
&lt;li&gt;Marketing &amp;amp; Consumer Research (Nielsen, Kantar Group, Ipsos, GfK, Euromonitor International)
&lt;ul&gt;
&lt;li&gt;Produce a consumer behavior analysis report for a leading retail brand. Identify key demographic shifts, purchasing trends, and brand loyalty factors, and provide actionable insights with data visualizations.&lt;/li&gt;
&lt;li&gt;Generate a detailed report on digital media consumption trends among millennials, incorporating survey results, social media analytics, and case studies of successful campaigns.&lt;/li&gt;
&lt;li&gt;Create a market segmentation report for a new consumer electronics launch. Identify key consumer segments, behavioral drivers, and media usage patterns with clear recommendations.&lt;/li&gt;
&lt;/ul&gt;
&lt;/li&gt;
&lt;li&gt;Financial Investment Research (Goldman Sachs, JPMorgan Chase, Morgan Stanley, Morningstar, Keefe Bruyette &amp;amp; Woods)
&lt;ul&gt;
&lt;li&gt;Produce an equity research report on mid-cap technology stocks. Include detailed financial modeling, valuation analysis, and buy/sell/hold recommendations with supporting data and charts.&lt;/li&gt;
&lt;li&gt;Generate a fixed income analysis report for corporate bonds in the industrial sector. Assess credit risk, yield forecasts, and macroeconomic influences, citing key data sources.&lt;/li&gt;
&lt;li&gt;Create a comprehensive report on global market trends impacting investment banking. Analyze regulatory changes, market sentiment, and performance metrics of leading financial institutions.&lt;/li&gt;
&lt;/ul&gt;
&lt;/li&gt;
&lt;li&gt;Healthcare Research (IQVIA, Frost &amp;amp; Sullivan, Evaluate Ltd, Deloitte Healthcare, IMS Health)
&lt;ul&gt;
&lt;li&gt;Produce a market analysis report on emerging biotechnologies in oncology. Include competitive landscape, regulatory challenges, and growth forecasts with relevant case studies.&lt;/li&gt;
&lt;li&gt;Generate a comprehensive report on patient satisfaction and telemedicine adoption trends. Analyze survey data from leading healthcare providers and benchmark best practices.&lt;/li&gt;
&lt;li&gt;Create a detailed study on pharmaceutical market dynamics in emerging economies. Focus on pipeline developments, regulatory environments, and market potential with actionable insights.&lt;/li&gt;
&lt;/ul&gt;
&lt;/li&gt;
&lt;li&gt;Legal Research Providers (LexisNexis, Westlaw, Bloomberg Law, Fastcase)
&lt;ul&gt;
&lt;li&gt;Produce a legal risk assessment report on the impact of recent data privacy regulations for multinational corporations. Include case studies, trend analysis (2019–2024), and strategic recommendations.&lt;/li&gt;
&lt;li&gt;Generate a comprehensive report summarizing key federal and Supreme Court rulings on intellectual property rights over the past five years, highlighting trends and divergent interpretations.&lt;/li&gt;
&lt;li&gt;Create a detailed report on the evolution of securities law and its effect on investment research practices, incorporating analysis of recent litigation and regulatory updates.&lt;/li&gt;
&lt;/ul&gt;
&lt;/li&gt;
&lt;li&gt;Media &amp;amp; News Research (Factiva, Kantar Media, Comscore, Cision)
&lt;ul&gt;
&lt;li&gt;Produce a media consumption trends report that analyzes audience behavior shifts across digital, TV, and print platforms. Include data visualizations, key drivers, and forecasted trends.&lt;/li&gt;
&lt;li&gt;Generate a comprehensive report on the impact of social media on traditional news reporting, with case studies and a comparative analysis of engagement metrics.&lt;/li&gt;
&lt;li&gt;Create a detailed study on the effectiveness of multimedia advertising campaigns, evaluating ROI, consumer engagement, and best practices with actionable insights.&lt;/li&gt;
&lt;/ul&gt;
&lt;/li&gt;
&lt;li&gt;Economic &amp;amp; Industry-Specific Research (Economist Intelligence Unit, BMI Research, IHS Markit, Consensus Economics)
&lt;ul&gt;
&lt;li&gt;Produce a macroeconomic outlook report for emerging markets, including GDP, inflation, and employment forecasts, with detailed data analysis and visualizations.&lt;/li&gt;
&lt;li&gt;Generate an industry analysis report on the automotive sector, covering technological innovations, competitive dynamics, and consolidation trends.&lt;/li&gt;
&lt;li&gt;Create a comprehensive country risk assessment report for a target region, detailing political, economic, and regulatory factors with recommendations for investors.&lt;/li&gt;
&lt;/ul&gt;
&lt;/li&gt;
&lt;li&gt;Human Resources &amp;amp; Employee Engagement Research (Gallup, Great Place to Work, Mercer)
&lt;ul&gt;
&lt;li&gt;Produce an employee engagement report for a multinational firm based on recent survey data. Identify key drivers of satisfaction, retention challenges, and improvement recommendations.&lt;/li&gt;
&lt;li&gt;Generate a comprehensive study on the impact of remote and hybrid work models on employee productivity across industries, including best practices and benchmark data.&lt;/li&gt;
&lt;li&gt;Create a detailed report on workplace culture transformation, analyzing organizational behavior trends, employee feedback, and actionable strategies to boost engagement.&lt;/li&gt;
&lt;/ul&gt;
&lt;/li&gt;
&lt;li&gt;Environmental, Social &amp;amp; Governance (ESG) Research (MSCI ESG Research, Sustainalytics, ISS ESG, Bloomberg ESG)
&lt;ul&gt;
&lt;li&gt;Produce an ESG performance report for a portfolio of global companies. Include sustainability scores, risk assessments, and recommendations for improvement with data visualizations.&lt;/li&gt;
&lt;li&gt;Generate a comprehensive study on the impact of climate change regulations on the energy sector, including policy analysis, market forecasts, and strategic implications.&lt;/li&gt;
&lt;li&gt;Create a detailed report on corporate social responsibility trends in the consumer goods industry, incorporating qualitative and quantitative analyses with actionable recommendations.&lt;/li&gt;
&lt;/ul&gt;
&lt;/li&gt;
&lt;li&gt;Education &amp;amp; Academic Research (RAND Corporation, National Center for Education Statistics, HolonIQ)
&lt;ul&gt;
&lt;li&gt;Produce an analysis report on the future of online education, examining technological adoption, market growth projections, and student outcome trends with supporting data.&lt;/li&gt;
&lt;li&gt;Generate a comprehensive study on the effects of educational policy reforms on public school performance in the U.S., including trend analysis and actionable recommendations.&lt;/li&gt;
&lt;li&gt;Create a detailed international higher education trends report, covering tuition dynamics, international student mobility, and emerging academic programs with comparative data.&lt;/li&gt;
&lt;/ul&gt;
&lt;/li&gt;
&lt;li&gt;Real Estate &amp;amp; Property Research (CBRE, JLL, CoStar Group, Cushman &amp;amp; Wakefield)
&lt;ul&gt;
&lt;li&gt;Produce a commercial real estate market analysis report for major urban centers, including occupancy trends, rental rate forecasts, and investment opportunity assessments.&lt;/li&gt;
&lt;li&gt;Generate a comprehensive study on residential housing market dynamics in emerging economies, focusing on affordability, supply-demand gaps, and policy impacts.&lt;/li&gt;
&lt;li&gt;Create a detailed report on the impact of urban redevelopment projects on local real estate values, including case studies, forecasts, and strategic recommendations.&lt;/li&gt;
&lt;/ul&gt;
&lt;/li&gt;
&lt;li&gt;Energy &amp;amp; Natural Resources Research (Wood Mackenzie, Rystad Energy, Bloomberg New Energy Finance)
&lt;ul&gt;
&lt;li&gt;Produce an analysis report on global renewable energy trends, covering technology adoption, market forecasts, and key policy drivers, with detailed data and visuals.&lt;/li&gt;
&lt;li&gt;Generate a comprehensive commodity price forecasting report for oil, natural gas, and key metals, incorporating historical trends, risk assessments, and predictive modeling.&lt;/li&gt;
&lt;li&gt;Create a detailed report on energy transition strategies for traditional energy companies, focusing on clean technology investments and market adaptation strategies.&lt;/li&gt;
&lt;/ul&gt;
&lt;/li&gt;
&lt;li&gt;Supply Chain &amp;amp; Logistics Research (ARC Advisory Group, Gartner Supply Chain Research, Supply Chain Insights)
&lt;ul&gt;
&lt;li&gt;Produce a report on supply chain resilience for global manufacturers. Analyze risk factors, digital transformation impacts, and best practices for operational efficiency with supporting data.&lt;/li&gt;
&lt;li&gt;Generate a comprehensive study on the impact of technology on logistics networks, including case studies on digital optimization and cost reduction strategies.&lt;/li&gt;
&lt;li&gt;Create a detailed report on emerging last-mile delivery solutions, assessing innovations, consumer expectations, and scalability with actionable insights.&lt;/li&gt;
&lt;/ul&gt;
&lt;/li&gt;
&lt;li&gt;Cybersecurity &amp;amp; Information Security Research (KuppingerCole, Forrester Security, IDC Cybersecurity, Cybersecurity Ventures)
&lt;ul&gt;
&lt;li&gt;Produce an in-depth report on emerging cybersecurity threats for large enterprises, including detailed analysis of recent incidents, risk vectors, and defense strategies.&lt;/li&gt;
&lt;li&gt;Generate a comprehensive cybersecurity market landscape report, evaluating vendor performance, technology forecasts, and best practices for mitigating risks.&lt;/li&gt;
&lt;li&gt;Create a detailed report on regulatory compliance trends in information security within the financial services industry, with case studies and strategic recommendations.&lt;/li&gt;
&lt;/ul&gt;
&lt;/li&gt;
&lt;li&gt;Social Media, Digital &amp;amp; Online Research (Comscore, SimilarWeb, Brandwatch)
&lt;ul&gt;
&lt;li&gt;Produce a digital audience behavior report for a global brand, focusing on social media trends, engagement metrics, and platform performance with detailed data analysis.&lt;/li&gt;
&lt;li&gt;Generate a comprehensive analysis of influencer marketing effectiveness across digital channels, including ROI metrics, case studies, and best practices.&lt;/li&gt;
&lt;li&gt;Create a detailed report on online brand sentiment analysis, incorporating social listening data, trend forecasts, and actionable recommendations.&lt;/li&gt;
&lt;/ul&gt;
&lt;/li&gt;
&lt;li&gt;Public Opinion &amp;amp; Political Research (Pew Research Center, Gallup, YouGov)
&lt;ul&gt;
&lt;li&gt;Produce a public opinion polling report on voter sentiment ahead of a major election. Include demographic breakdowns, key issue analysis, and trend visualizations for the past five years.&lt;/li&gt;
&lt;li&gt;Generate a comprehensive study on political risk in emerging markets, analyzing historical data, current trends, and future projections, with policy recommendations.&lt;/li&gt;
&lt;li&gt;Create a detailed report on the influence of media on public policy, using survey data, social media analysis, and comparative case studies.&lt;/li&gt;
&lt;/ul&gt;
&lt;/li&gt;
&lt;li&gt;Sports, Entertainment &amp;amp; Media Research (Nielsen Sports, Sportcal, Kantar Media Sports)
&lt;ul&gt;
&lt;li&gt;Produce a market analysis report on sports sponsorship trends, detailing viewership metrics, brand engagement, and investment ROI with industry case studies.&lt;/li&gt;
&lt;li&gt;Generate a comprehensive report on audience behavior in the streaming media industry, including demographic insights, consumption trends, and competitive benchmarks.&lt;/li&gt;
&lt;li&gt;Create a detailed analysis of digital advertising effectiveness in the entertainment sector, including segmentation data, ROI analysis, and strategic recommendations.&lt;/li&gt;
&lt;/ul&gt;
&lt;/li&gt;
&lt;li&gt;Innovation, R&amp;amp;D &amp;amp; Technology Trends Research (Innosight, Frost &amp;amp; Sullivan Innovation, CB Insights)
&lt;ul&gt;
&lt;li&gt;Produce a global R&amp;amp;D investment trends report, analyzing technology spending, innovation indices, and the impact on market growth across key industries.&lt;/li&gt;
&lt;li&gt;Generate a comprehensive study on disruptive technologies in manufacturing, including competitive analysis, market potential forecasts, and adoption trends.&lt;/li&gt;
&lt;li&gt;Create a detailed report on emerging innovation hubs worldwide, focusing on startup ecosystems, funding trends, and collaborative opportunities in technology.&lt;/li&gt;
&lt;/ul&gt;
&lt;/li&gt;
&lt;li&gt;Agriculture &amp;amp; Agribusiness Research (Rabobank Agribusiness Research, USDA Economic Research Service, AgFunder)
&lt;ul&gt;
&lt;li&gt;Produce an analysis report on global agricultural market trends, including crop yield forecasts, trade dynamics, and policy impacts, with data visualizations.&lt;/li&gt;
&lt;li&gt;Generate a comprehensive study on agritech innovations such as precision farming and sustainable practices, including case studies and market forecasts.&lt;/li&gt;
&lt;li&gt;Create a detailed report on the impact of climate change on food production and supply chain stability in agribusiness, with risk assessments and strategic recommendations.&lt;/li&gt;
&lt;/ul&gt;
&lt;/li&gt;
&lt;li&gt;Environmental &amp;amp; Climate Change Research (Carbon Trust, IHS Markit Energy Transition, Bloomberg New Energy Finance)
&lt;ul&gt;
&lt;li&gt;Produce a report on the economic and social impacts of climate change on urban infrastructure, including forecasting models and policy recommendations.&lt;/li&gt;
&lt;li&gt;Generate a comprehensive study on national climate policies and their effects on industrial competitiveness, with detailed trend analysis and source citations.&lt;/li&gt;
&lt;li&gt;Create a detailed report on corporate sustainability initiatives, assessing environmental risk management practices and providing actionable recommendations for improvement.&lt;/li&gt;
&lt;/ul&gt;
&lt;/li&gt;
&lt;li&gt;Customer Experience (CX) &amp;amp; User Experience (UX) Research (Forrester CX Research, Gartner CX Research, Qualtrics, Nielsen Norman Group)
&lt;ul&gt;
&lt;li&gt;Produce a report on customer journey mapping for a leading retail brand, identifying key touchpoints, pain points, and actionable improvement strategies with data visualizations.&lt;/li&gt;
&lt;li&gt;Generate a comprehensive study on digital user experience trends for e-commerce platforms, including usability testing insights, design best practices, and conversion optimization recommendations.&lt;/li&gt;
&lt;li&gt;Create a detailed report on customer satisfaction and loyalty metrics across multiple industries, integrating survey data and actionable recommendations to enhance overall CX.&lt;/li&gt;
&lt;/ul&gt;
&lt;/li&gt;
&lt;li&gt;Blockchain, Cryptocurrency &amp;amp; Fintech Research (Chainalysis, CoinDesk Research, Deloitte Fintech Research, CB Insights)
&lt;ul&gt;
&lt;li&gt;Produce an analysis report on emerging blockchain technologies and their applications in financial services, including market trends, adoption forecasts, and case studies.&lt;/li&gt;
&lt;li&gt;Generate a comprehensive study on cryptocurrency market dynamics, analyzing regulatory developments, investor sentiment, and competitive landscapes with source citations.&lt;/li&gt;
&lt;li&gt;Create a detailed report on fintech disruption in traditional banking, with case studies on leading startups, technology adoption, and future market forecasts.&lt;/li&gt;
&lt;/ul&gt;
&lt;/li&gt;
&lt;li&gt;Venture Capital, Startup &amp;amp; Private Equity Research (PitchBook, CB Insights, Crunchbase, Preqin)
&lt;ul&gt;
&lt;li&gt;Produce a global venture capital investment trends report, including performance analysis of high-growth startups, sector benchmarks, and emerging market opportunities.&lt;/li&gt;
&lt;li&gt;Generate a comprehensive study on private equity market dynamics, covering deal flow analysis, exit strategies, and forecasted trends with supporting data.&lt;/li&gt;
&lt;li&gt;Create a detailed report on emerging startup ecosystems in key regions, highlighting funding trends, investor activity, and growth potential with actionable insights.&lt;/li&gt;
&lt;/ul&gt;
&lt;/li&gt;
&lt;li&gt;Operations Research &amp;amp; Management Science Consulting (The Brattle Group, NERA Economic Consulting, CRA International)
&lt;ul&gt;
&lt;li&gt;Produce a report on optimization techniques for operational efficiency in large-scale manufacturing, including quantitative analysis, simulation models, and case studies.&lt;/li&gt;
&lt;li&gt;Generate a comprehensive study on the application of predictive analytics in supply chain management, focusing on data modeling, process improvements, and actionable insights.&lt;/li&gt;
&lt;li&gt;Create a detailed report on advanced quantitative modeling approaches to solve complex business problems in logistics and operations, including scenario analysis and recommendations.&lt;/li&gt;
&lt;/ul&gt;
&lt;/li&gt;
&lt;li&gt;Cultural &amp;amp; Social Research (Ethnographic/Sociocultural Studies) (Ipsos MORI, Kantar TNS, YouGov)
&lt;ul&gt;
&lt;li&gt;Produce a qualitative ethnographic study on urban consumer lifestyle trends, incorporating field observations, interviews, and cultural analysis with actionable insights.&lt;/li&gt;
&lt;li&gt;Generate a comprehensive study on how cultural shifts influence global brand perception, including comparative case studies and trend analysis.&lt;/li&gt;
&lt;li&gt;Create a detailed report on sociocultural dynamics and consumer behavior in emerging economies, integrating in-depth field research and actionable recommendations.&lt;/li&gt;
&lt;/ul&gt;
&lt;/li&gt;
&lt;li&gt;Economic &amp;amp; Demographic Research Firms (Oxford Economics, The Conference Board, CEIC Data)
&lt;ul&gt;
&lt;li&gt;Produce a macroeconomic forecasting report for a specific region, including GDP, inflation, and employment trends with detailed data visualizations and source citations.&lt;/li&gt;
&lt;li&gt;Generate a detailed demographic analysis report for a target market, highlighting age distribution, income levels, and consumption patterns with actionable insights.&lt;/li&gt;
&lt;li&gt;Create a comprehensive report on the economic impact of demographic shifts on consumer markets, with policy recommendations and trend analysis.&lt;/li&gt;
&lt;/ul&gt;
&lt;/li&gt;
&lt;li&gt;Academic &amp;amp; Think Tank Research Organizations (Brookings Institution, RAND Corporation, Carnegie Endowment for International Peace)
&lt;ul&gt;
&lt;li&gt;Produce a policy research report on global governance challenges and their implications for economic development, including case studies, literature reviews, and expert interviews.&lt;/li&gt;
&lt;li&gt;Generate a comprehensive study on social inequality and its effects on public health and education outcomes, supported by empirical research and trend analysis.&lt;/li&gt;
&lt;li&gt;Create a detailed report on emerging trends in international relations and their impact on global trade and security, integrating academic research and data analytics.&lt;/li&gt;
&lt;/ul&gt;
&lt;/li&gt;
&lt;li&gt;Market Research Technology &amp;amp; Software Providers (Qualtrics, SurveyMonkey, Confirmit)
&lt;ul&gt;
&lt;li&gt;Produce a report on the latest innovations in survey technology and data analytics software for market research, including product comparisons, user case studies, and future trend forecasts.&lt;/li&gt;
&lt;li&gt;Generate a comprehensive study on the integration of AI and machine learning in consumer insights platforms, highlighting case studies, performance metrics, and industry benchmarks.&lt;/li&gt;
&lt;li&gt;Create a detailed report on digital transformation trends in market research technology, featuring analysis of leading software solutions, market share data, and recommendations for technology adoption.&lt;/li&gt;
&lt;/ul&gt;
&lt;/li&gt;
&lt;/ol&gt;
&lt;/li&gt;
&lt;li&gt;When evaluating inputs, models tend to prefer the first response, prefer their own response, and prefer longer responses. &lt;a href=&#34;https://sub.thursdai.news/p/thursdai-feb-20-live-from-ai-eng&#34;&gt;ThursdAI&lt;/a&gt;&lt;/li&gt;
&lt;li&gt;Real-time speech-to-text options for transcription:
&lt;ul&gt;
&lt;li&gt;&lt;a href=&#34;https://deepgram.com/learn/live-transcription-mic-browser&#34;&gt;Deepgram&lt;/a&gt; has a MediaRecorder API, which is perfect.&lt;/li&gt;
&lt;li&gt;&lt;a href=&#34;https://github.com/QuentinFuxa/whisper_streaming_web&#34;&gt;Whisper Streaming Web&lt;/a&gt; is a web app that can transcribe audio real-time from the browser. A good approach, but I wouldn&amp;rsquo;t use it for meeting transcription on my mid-end laptop. Streaming takes up the bulk of my GPU, leaving little for transcription.&lt;/li&gt;
&lt;li&gt;&lt;a href=&#34;https://github.com/collabora/WhisperLive&#34;&gt;whisper-live&lt;/a&gt; runs as a Python console app and does something similar.&lt;/li&gt;
&lt;li&gt;&lt;a href=&#34;https://huggingface.co/spaces/Xenova/realtime-whisper-webgpu&#34;&gt;Whisper WebGPU&lt;/a&gt; runs on the browser (only 200MB). Cool! But slow and still takes up GPU.&lt;/li&gt;
&lt;/ul&gt;
&lt;/li&gt;
&lt;li&gt;&lt;a href=&#34;https://github.com/gpt-omni/mini-omni/&#34;&gt;Mini-omni&lt;/a&gt; is an open-source Qwen-based LLM that can hear and talk while thinking in real-time. An interesting experiment, but not for prototyping.&lt;/li&gt;
&lt;li&gt;OpenAI shares an insights report with clients that has insights on what different professions search for. What doctors search for is:
&lt;ol&gt;
&lt;li&gt;Is my diagnosis right?&lt;/li&gt;
&lt;li&gt;How do I read this report?&lt;/li&gt;
&lt;li&gt;Is my prescription correct?&lt;/li&gt;
&lt;li&gt;Is there a cheaper medicine?&lt;/li&gt;
&lt;li&gt;What&amp;rsquo;s the life expectancy given these symptoms?&lt;/li&gt;
&lt;/ol&gt;
&lt;/li&gt;
&lt;li&gt;Dataclasses in Python have a slight overhead over named tuples. The 2 main uses I see for them are: providing defaults and offering type hints.&lt;/li&gt;
&lt;li&gt;UVB 76 is a radio channel has been broadcasting static (with occasional Russian conversation) since 1976. No one knows why. It&amp;rsquo;s live at &lt;a href=&#34;https://m.youtube.com/watch?v=8h_D2P0iqMk&#34;&gt;https://m.youtube.com/watch?v=8h_D2P0iqMk&lt;/a&gt;&lt;/li&gt;
&lt;li&gt;Romans washed clothes in urine. The government taxed the purchase of urine for commercial purposes! That&amp;rsquo;s the origin of the phrase &amp;ldquo;Pecunia non olet&amp;rdquo; which means &amp;ldquo;money doesn&amp;rsquo;t stink&amp;rdquo;.&lt;/li&gt;
&lt;li&gt;&lt;a href=&#34;https://nixos.org/&#34;&gt;Nix&lt;/a&gt; is a package manager that creates container-like environments. Like a cross between Docker and &lt;code&gt;apt&lt;/code&gt; / &lt;code&gt;venv&lt;/code&gt;. It has an immutable file system. &lt;a href=&#34;https://www.jetify.com/devbox&#34;&gt;DevBox&lt;/a&gt; is a higher-level tool built on top of Nix that streamlines developer workflows, e.g. common project environment setup.&lt;/li&gt;
&lt;li&gt;VS Code can be used to develop inside a Docker container via Podman, too. Set &lt;code&gt;dev.containers.dockerPath&amp;quot;: &amp;quot;podman&amp;quot;&lt;/code&gt; &lt;a href=&#34;https://geekingoutpodcast.substack.com/p/running-dev-containers-locally-with&#34;&gt;Ref&lt;/a&gt;&lt;/li&gt;
&lt;li&gt;&lt;a href=&#34;https://www.rilldata.com/&#34;&gt;Rill Data&lt;/a&gt; is an interesting BI tool based on DuckDB. It auto-generates a dashboard given a dataset.&lt;/li&gt;
&lt;li&gt;It&amp;rsquo;s possible to assign &amp;ldquo;variables&amp;rdquo; in SQL (notably in DuckDB). Here&amp;rsquo;s an example:
&lt;div class=&#34;highlight&#34;&gt;&lt;pre tabindex=&#34;0&#34; class=&#34;chroma&#34;&gt;&lt;code class=&#34;language-sql&#34; data-lang=&#34;sql&#34;&gt;&lt;span class=&#34;line&#34;&gt;&lt;span class=&#34;cl&#34;&gt;&lt;span class=&#34;k&#34;&gt;WITH&lt;/span&gt;&lt;span class=&#34;w&#34;&gt;
&lt;/span&gt;&lt;/span&gt;&lt;/span&gt;&lt;span class=&#34;line&#34;&gt;&lt;span class=&#34;cl&#34;&gt;&lt;span class=&#34;w&#34;&gt;  &lt;/span&gt;&lt;span class=&#34;n&#34;&gt;sessions&lt;/span&gt;&lt;span class=&#34;w&#34;&gt; &lt;/span&gt;&lt;span class=&#34;k&#34;&gt;AS&lt;/span&gt;&lt;span class=&#34;w&#34;&gt; &lt;/span&gt;&lt;span class=&#34;p&#34;&gt;(&lt;/span&gt;&lt;span class=&#34;k&#34;&gt;FROM&lt;/span&gt;&lt;span class=&#34;w&#34;&gt; &lt;/span&gt;&lt;span class=&#34;n&#34;&gt;events&lt;/span&gt;&lt;span class=&#34;w&#34;&gt; &lt;/span&gt;&lt;span class=&#34;k&#34;&gt;SELECT&lt;/span&gt;&lt;span class=&#34;w&#34;&gt; &lt;/span&gt;&lt;span class=&#34;k&#34;&gt;COUNT&lt;/span&gt;&lt;span class=&#34;p&#34;&gt;(&lt;/span&gt;&lt;span class=&#34;k&#34;&gt;DISTINCT&lt;/span&gt;&lt;span class=&#34;w&#34;&gt; &lt;/span&gt;&lt;span class=&#34;n&#34;&gt;session_id&lt;/span&gt;&lt;span class=&#34;p&#34;&gt;)&lt;/span&gt;&lt;span class=&#34;w&#34;&gt; &lt;/span&gt;&lt;span class=&#34;k&#34;&gt;AS&lt;/span&gt;&lt;span class=&#34;w&#34;&gt; &lt;/span&gt;&lt;span class=&#34;n&#34;&gt;value&lt;/span&gt;&lt;span class=&#34;p&#34;&gt;),&lt;/span&gt;&lt;span class=&#34;w&#34;&gt;
&lt;/span&gt;&lt;/span&gt;&lt;/span&gt;&lt;span class=&#34;line&#34;&gt;&lt;span class=&#34;cl&#34;&gt;&lt;span class=&#34;w&#34;&gt;  &lt;/span&gt;&lt;span class=&#34;n&#34;&gt;pages&lt;/span&gt;&lt;span class=&#34;w&#34;&gt; &lt;/span&gt;&lt;span class=&#34;k&#34;&gt;AS&lt;/span&gt;&lt;span class=&#34;w&#34;&gt; &lt;/span&gt;&lt;span class=&#34;p&#34;&gt;(&lt;/span&gt;&lt;span class=&#34;k&#34;&gt;FROM&lt;/span&gt;&lt;span class=&#34;w&#34;&gt; &lt;/span&gt;&lt;span class=&#34;n&#34;&gt;events&lt;/span&gt;&lt;span class=&#34;w&#34;&gt; &lt;/span&gt;&lt;span class=&#34;k&#34;&gt;SELECT&lt;/span&gt;&lt;span class=&#34;w&#34;&gt; &lt;/span&gt;&lt;span class=&#34;k&#34;&gt;COUNT&lt;/span&gt;&lt;span class=&#34;p&#34;&gt;(&lt;/span&gt;&lt;span class=&#34;o&#34;&gt;*&lt;/span&gt;&lt;span class=&#34;p&#34;&gt;)&lt;/span&gt;&lt;span class=&#34;w&#34;&gt; &lt;/span&gt;&lt;span class=&#34;k&#34;&gt;AS&lt;/span&gt;&lt;span class=&#34;w&#34;&gt; &lt;/span&gt;&lt;span class=&#34;n&#34;&gt;value&lt;/span&gt;&lt;span class=&#34;p&#34;&gt;)&lt;/span&gt;&lt;span class=&#34;w&#34;&gt;
&lt;/span&gt;&lt;/span&gt;&lt;/span&gt;&lt;span class=&#34;line&#34;&gt;&lt;span class=&#34;cl&#34;&gt;&lt;span class=&#34;k&#34;&gt;FROM&lt;/span&gt;&lt;span class=&#34;w&#34;&gt; &lt;/span&gt;&lt;span class=&#34;n&#34;&gt;sessions&lt;/span&gt;&lt;span class=&#34;p&#34;&gt;,&lt;/span&gt;&lt;span class=&#34;w&#34;&gt; &lt;/span&gt;&lt;span class=&#34;n&#34;&gt;pages&lt;/span&gt;&lt;span class=&#34;w&#34;&gt;
&lt;/span&gt;&lt;/span&gt;&lt;/span&gt;&lt;span class=&#34;line&#34;&gt;&lt;span class=&#34;cl&#34;&gt;&lt;span class=&#34;k&#34;&gt;SELECT&lt;/span&gt;&lt;span class=&#34;w&#34;&gt; &lt;/span&gt;&lt;span class=&#34;n&#34;&gt;sessions&lt;/span&gt;&lt;span class=&#34;p&#34;&gt;.&lt;/span&gt;&lt;span class=&#34;n&#34;&gt;value&lt;/span&gt;&lt;span class=&#34;w&#34;&gt; &lt;/span&gt;&lt;span class=&#34;o&#34;&gt;/&lt;/span&gt;&lt;span class=&#34;w&#34;&gt; &lt;/span&gt;&lt;span class=&#34;n&#34;&gt;pages&lt;/span&gt;&lt;span class=&#34;p&#34;&gt;.&lt;/span&gt;&lt;span class=&#34;n&#34;&gt;value&lt;/span&gt;&lt;span class=&#34;w&#34;&gt; &lt;/span&gt;&lt;span class=&#34;k&#34;&gt;AS&lt;/span&gt;&lt;span class=&#34;w&#34;&gt; &lt;/span&gt;&lt;span class=&#34;n&#34;&gt;pages_per_session&lt;/span&gt;&lt;span class=&#34;p&#34;&gt;;&lt;/span&gt;&lt;span class=&#34;w&#34;&gt;
&lt;/span&gt;&lt;/span&gt;&lt;/span&gt;&lt;/code&gt;&lt;/pre&gt;&lt;/div&gt;&lt;/li&gt;
&lt;li&gt;DuckDB has a &lt;code&gt;GROUP BY *&lt;/code&gt; that groups by all categorical columns. &lt;code&gt;SELECT x, y, COUNT(*) FROM t GROUP BY *&lt;/code&gt; is equivalent to &lt;code&gt;SELECT x, y, COUNT(*) FROM t GROUP BY x, y&lt;/code&gt;.&lt;/li&gt;
&lt;li&gt;VS Code can be used as a code executor by adding &lt;code&gt;{&amp;quot;key&amp;quot;: &amp;quot;shift+enter&amp;quot;, &amp;quot;command&amp;quot;: &amp;quot;workbench.action.terminal.runSelectedText&amp;quot;, &amp;quot;when&amp;quot;: &amp;quot;editorFocus&amp;quot;}&lt;/code&gt; to the &lt;code&gt;keybindings.json&lt;/code&gt; file. Press Shift-Enter to run the selection on the terminal. Useful for DuckDB, SQLite, etc. &lt;a href=&#34;https://motherduck.com/blog/duckdb-tutorial-for-beginners/&#34;&gt;Ref&lt;/a&gt;&lt;/li&gt;
&lt;li&gt;LLMs are excellent at database migration. They can convert schemas and queries across SQL dialects (e.g. BigQuery to DuckDB, etc.) at 90%+ accuracy. This is useful when clients want to migrate cloud providers, go from on-prem to cloud, or reduce cost by switching databases.&lt;/li&gt;
&lt;/ul&gt;
</description>
    </item>
    <item>
      <title>Things I Learned - 17 Nov 2024</title>
      <link>https://www.s-anand.net/blog/things-i-learned-17-nov-2024/</link>
      <pubDate>Sun, 17 Nov 2024 00:00:00 +0000</pubDate>
      <guid>https://www.s-anand.net/blog/things-i-learned-17-nov-2024/</guid>
      <description>&lt;p&gt;This week, I learned:&lt;/p&gt;
&lt;ul&gt;
&lt;li&gt;Anthropic has single-plage docs for LLMs. &lt;a href=&#34;https://docs.anthropic.com/llms.txt&#34;&gt;Condensed version&lt;/a&gt; and &lt;a href=&#34;https://docs.anthropic.com/llms-full.txt&#34;&gt;Full version&lt;/a&gt;&lt;/li&gt;
&lt;li&gt;&lt;a href=&#34;https://www.ted.com/pages/malcolm-gladwell-on-the-importance-of-self-correction-transcript&#34;&gt;Malcolm Gladwell on the importance of self-correction&lt;/a&gt;
&lt;ul&gt;
&lt;li&gt;Belonging to multiple social worlds is a good way to defend against no longer being good at what you used to be. Diverse values and social groups help.&lt;/li&gt;
&lt;li&gt;Self handicapping explains a lot about the world. You study late for a maths test - so you can fail for lack of trying, not aptitude. Ecosystems (e.g. sports teams) mitigate self-handicapping.&lt;/li&gt;
&lt;li&gt;You don&amp;rsquo;t have to be good in athletics to get the benefits. A slow runner gets the same discipline, pumping up, etc that a fast runner does&lt;/li&gt;
&lt;li&gt;Mono cultures are good to accomplish a known mission. Diversity is good to pivot during uncertainty. So, localize mono cultures&lt;/li&gt;
&lt;li&gt;Diversity helps only if there are sufficient numbers, or if they have enough power to change the organization&amp;rsquo;s thinking.&lt;/li&gt;
&lt;/ul&gt;
&lt;/li&gt;
&lt;li&gt;Use a standardized password strategy, e.g. use the month like GramNov2024 (via Namit)&lt;/li&gt;
&lt;li&gt;Gemini has an OpenAI compatible API. &lt;a href=&#34;https://ai.google.dev/gemini-api/docs/openai&#34;&gt;Gemini Docs&lt;/a&gt;&lt;/li&gt;
&lt;li&gt;Ethan Mollick says Claude is solving MBA case studies well. &lt;a href=&#34;https://x.com/emollick/status/1856161026238025835&#34;&gt;x.com&lt;/a&gt;&lt;/li&gt;
&lt;li&gt;LLMs pay a lot of attention to the first 6 tokens. &lt;a href=&#34;https://huggingface.co/blog/tomaarsen/attention-sinks&#34;&gt;Ref&lt;/a&gt;&lt;/li&gt;
&lt;li&gt;This is an interesting article on &amp;ldquo;UI in the age of Gen AI&amp;rdquo;. &lt;a href=&#34;https://agao.substack.com/p/uiux-in-the-age-of-generative-ai&#34;&gt;Ref&lt;/a&gt;&lt;/li&gt;
&lt;li&gt;Google Open sourced Alphafold 3. &lt;a href=&#34;https://github.com/google-deepmind/alphafold3&#34;&gt;Repo&lt;/a&gt;&lt;/li&gt;
&lt;li&gt;Cloudflare R2 has the same API as S3 but is cheaper&lt;/li&gt;
&lt;li&gt;Prefect.io is a good alternative to Airflow / cron. Can use for synchronisation tasks, e.g. Drive to server. But no Auth, UI params or config.&lt;/li&gt;
&lt;li&gt;Gemini transcription does not give accurate timestamps. Whisper does. But the quality of transcription is similar.&lt;/li&gt;
&lt;li&gt;Pass a complex data structure to Claude.ai and have it create an app to visualize it. It does well. &lt;a href=&#34;https://x.com/simonw/status/1855819673482461216&#34;&gt;Simin Willison&lt;/a&gt;&lt;/li&gt;
&lt;li&gt;&lt;a href=&#34;https://techcouncilventures.com/&#34;&gt;Tech Council Ventures&lt;/a&gt; and &lt;a href=&#34;https://sunicon.vc/&#34;&gt;Sunicon VC&lt;/a&gt; invest in early stage startups, and aloso provide them technology support (via Naveen)&lt;/li&gt;
&lt;/ul&gt;
</description>
    </item>
    <item>
      <title>Tools to publish annotated talks from videos</title>
      <link>https://www.s-anand.net/blog/tools-to-publish-annotated-talks-from-videos/</link>
      <pubDate>Sun, 20 Oct 2024 07:22:57 +0000</pubDate>
      <guid>https://www.s-anand.net/blog/tools-to-publish-annotated-talks-from-videos/</guid>
      <description>&lt;p&gt;&lt;img alt=&#34;Tools to publish annotated talks from videos&#34; loading=&#34;lazy&#34; src=&#34;https://www.s-anand.net/blog/assets/maxresdefault.webp&#34;&gt;&lt;/p&gt;
&lt;p&gt;&lt;a href=&#34;https://www.linkedin.com/in/arun-tangirala-1712444/&#34;&gt;Arun Tangirala&lt;/a&gt; and I webinared on &amp;ldquo;AI in Education&amp;rdquo; yesterday.&lt;/p&gt;
&lt;p&gt;&lt;img alt=&#34;(PS: &amp;quot;Webinared&amp;quot; is not a word. But &amp;quot;verbing weirds language&amp;quot;.)&#34; loading=&#34;lazy&#34; src=&#34;https://picayune.uclick.com/comics/ch/1993/ch930125.gif&#34;&gt;&lt;/p&gt;
&lt;p&gt;This post isn&amp;rsquo;t about the webinar, which went on for an hour and was good fun.&lt;/p&gt;
&lt;div class=&#34;video-embed&#34;&gt;&lt;iframe src=&#34;https://www.youtube.com/embed/fPvhDOyUPc8&#34; title=&#34;YouTube video&#34; loading=&#34;lazy&#34; allow=&#34;accelerometer; autoplay; clipboard-write; encrypted-media; gyroscope; picture-in-picture&#34; allowfullscreen&gt;&lt;/iframe&gt;&lt;/div&gt;
&lt;p&gt;This post isn&amp;rsquo;t for my preparation for the webinar, which happened frantically 15 minutes before it started.&lt;/p&gt;
&lt;p&gt;This post is about how I created the annotated talk at &lt;a href=&#34;https://github.com/sanand0/ai-in-education-webinar&#34;&gt;https://github.com/sanand0/ai-in-education-webinar&lt;/a&gt; (inspired by &lt;a href=&#34;https://simonwillison.net/2023/Aug/6/annotated-presentations/&#34;&gt;Simon Willison&amp;rsquo;s annotated presentations&lt;/a&gt; process) &amp;ndash; a post-processing step that took ~3 hours &amp;ndash; and the tools I used for this.&lt;/p&gt;
&lt;h4 id=&#34;scrape-the-comments&#34;&gt;Scrape the comments&lt;/h4&gt;
&lt;p&gt;The Hindu used &lt;a href=&#34;https://streamyard.com/&#34;&gt;StreamYard&lt;/a&gt;. It web-based and has a comments section. I used JS in the &lt;a href=&#34;https://developer.chrome.com/docs/devtools/console&#34;&gt;DevTools Console&lt;/a&gt; to scrape. Roughly, &lt;code&gt;$$(&amp;quot;.some-class-name&amp;quot;).map(d =&amp;gt; d.textContent)&lt;/code&gt;&lt;/p&gt;
&lt;p&gt;But the comments are not all visible together. As you scroll, newer/older comments are loaded. So I needed to use my favorite technique: &lt;a href=&#34;https://www.s-anand.net/blog/cyborg-scraping/&#34;&gt;Cyborg Scraping&lt;/a&gt;. During Q&amp;amp;A, I kept scrolling to the bottom and ran:&lt;/p&gt;
&lt;div class=&#34;highlight&#34;&gt;&lt;pre tabindex=&#34;0&#34; class=&#34;chroma&#34;&gt;&lt;code class=&#34;language-javascript&#34; data-lang=&#34;javascript&#34;&gt;&lt;span class=&#34;line&#34;&gt;&lt;span class=&#34;cl&#34;&gt;&lt;span class=&#34;c1&#34;&gt;// One-time set-up
&lt;/span&gt;&lt;/span&gt;&lt;/span&gt;&lt;span class=&#34;line&#34;&gt;&lt;span class=&#34;cl&#34;&gt;&lt;span class=&#34;nx&#34;&gt;messages&lt;/span&gt; &lt;span class=&#34;o&#34;&gt;=&lt;/span&gt; &lt;span class=&#34;k&#34;&gt;new&lt;/span&gt; &lt;span class=&#34;nx&#34;&gt;Set&lt;/span&gt;&lt;span class=&#34;p&#34;&gt;();&lt;/span&gt;
&lt;/span&gt;&lt;/span&gt;&lt;span class=&#34;line&#34;&gt;&lt;span class=&#34;cl&#34;&gt;
&lt;/span&gt;&lt;/span&gt;&lt;span class=&#34;line&#34;&gt;&lt;span class=&#34;cl&#34;&gt;&lt;span class=&#34;c1&#34;&gt;// Run every now and then after scrolling to the bottom
&lt;/span&gt;&lt;/span&gt;&lt;/span&gt;&lt;span class=&#34;line&#34;&gt;&lt;span class=&#34;cl&#34;&gt;&lt;span class=&#34;c1&#34;&gt;// Stores all messages without duplication
&lt;/span&gt;&lt;/span&gt;&lt;/span&gt;&lt;span class=&#34;line&#34;&gt;&lt;span class=&#34;cl&#34;&gt;&lt;span class=&#34;nx&#34;&gt;$$&lt;/span&gt;&lt;span class=&#34;p&#34;&gt;(&lt;/span&gt;&lt;span class=&#34;s2&#34;&gt;&amp;#34;.some-class-name&amp;#34;&lt;/span&gt;&lt;span class=&#34;p&#34;&gt;).&lt;/span&gt;&lt;span class=&#34;nx&#34;&gt;map&lt;/span&gt;&lt;span class=&#34;p&#34;&gt;(&lt;/span&gt;&lt;span class=&#34;nx&#34;&gt;d&lt;/span&gt; &lt;span class=&#34;p&#34;&gt;=&amp;gt;&lt;/span&gt; &lt;span class=&#34;nx&#34;&gt;messages&lt;/span&gt;&lt;span class=&#34;p&#34;&gt;.&lt;/span&gt;&lt;span class=&#34;nx&#34;&gt;add&lt;/span&gt;&lt;span class=&#34;p&#34;&gt;(&lt;/span&gt;&lt;span class=&#34;nx&#34;&gt;d&lt;/span&gt;&lt;span class=&#34;p&#34;&gt;.&lt;/span&gt;&lt;span class=&#34;nx&#34;&gt;textContent&lt;/span&gt;&lt;span class=&#34;p&#34;&gt;));&lt;/span&gt;
&lt;/span&gt;&lt;/span&gt;&lt;span class=&#34;line&#34;&gt;&lt;span class=&#34;cl&#34;&gt;
&lt;/span&gt;&lt;/span&gt;&lt;span class=&#34;line&#34;&gt;&lt;span class=&#34;cl&#34;&gt;&lt;span class=&#34;c1&#34;&gt;// Finally, copy the messages as a JSON array to the clipboard
&lt;/span&gt;&lt;/span&gt;&lt;/span&gt;&lt;span class=&#34;line&#34;&gt;&lt;span class=&#34;cl&#34;&gt;&lt;span class=&#34;nx&#34;&gt;copy&lt;/span&gt;&lt;span class=&#34;p&#34;&gt;([...&lt;/span&gt;&lt;span class=&#34;nx&#34;&gt;messages&lt;/span&gt;&lt;span class=&#34;p&#34;&gt;]);&lt;/span&gt;
&lt;/span&gt;&lt;/span&gt;&lt;/code&gt;&lt;/pre&gt;&lt;/div&gt;&lt;p&gt;I used VS Code&amp;rsquo;s regular expression search &lt;code&gt;^\d\d:\d\d (AM|PM)$&lt;/code&gt; to find the timestamps and split the name, time, and comments into columns. &lt;a href=&#34;https://code.visualstudio.com/docs/editor/codebasics#_multiple-selections-multicursor&#34;&gt;Multiple-cursors&lt;/a&gt; all the way. Then I pasted it in Excel to convert it to Markdown. I added this in the &lt;a href=&#34;https://github.com/sanand0/ai-in-education-webinar?tab=readme-ov-file#comments&#34;&gt;Comments in the Chat&lt;/a&gt; section.&lt;/p&gt;
&lt;p&gt;(&lt;strong&gt;Excel&lt;/strong&gt; to convert to Markdown? Yeah. My formula is below.)&lt;/p&gt;
&lt;p&gt;&lt;img loading=&#34;lazy&#34; src=&#34;https://www.s-anand.net/blog/assets/excel-comments.webp&#34;&gt;&lt;/p&gt;
&lt;h4 id=&#34;transcribe-the-video&#34;&gt;Transcribe the video&lt;/h4&gt;
&lt;p&gt;I downloaded &lt;a href=&#34;https://youtu.be/fPvhDOyUPc8&#34;&gt;the video&lt;/a&gt; using &lt;a href=&#34;https://github.com/yt-dlp/yt-dlp&#34;&gt;yt-dlp&lt;/a&gt;, which I find the most robust tool for YouTube downloads.&lt;/p&gt;
&lt;p&gt;I used &lt;code&gt;ffmpeg.exe -i webinar.mp4 -b:a 32k -ac 1 -ar 22050 webinar.mp3&lt;/code&gt; to convert the video to audio. I use these settings for voice (not music) to get a fairly small MP3 file. I should have used Opus, which is much smaller. I&amp;rsquo;ll do that next.)&lt;/p&gt;
&lt;p&gt;Groq recently added &lt;a href=&#34;https://groq.com/whisper-large-v3-turbo-now-available-on-groq-combining-speed-quality-for-speech-recognition/&#34;&gt;Whisper Large v3&lt;/a&gt; (which is better than most earlier models on transcription.) So I could just go to the &lt;a href=&#34;https://console.groq.com/playground&#34;&gt;Groq playground&lt;/a&gt; and upload the MP3 file to get a transcript in a few seconds.&lt;/p&gt;
&lt;h4 id=&#34;add-images-to-the-transcript&#34;&gt;Add images to the transcript&lt;/h4&gt;
&lt;p&gt;I wrote a tool, &lt;a href=&#34;https://github.com/gramener/videoscribe&#34;&gt;VideoScribe&lt;/a&gt; (WIP), to make transcription and image insertion easy. It uses &lt;code&gt;ffmpeg -i webinar.mp4 -vf select=&#39;key&#39;,showinfo -vsync vfr -compression_level 10 &amp;quot;%04d.jpg&amp;quot;&lt;/code&gt; to extract all keyframes (images with major changes) from the video and inserts them into the right spots in the transcript.&lt;/p&gt;
&lt;p&gt;I picked 36 out of the ~700 that were generated as representing new slides, questions, or key moments and exported it as Markdown. I also used VS Code &lt;a href=&#34;https://code.visualstudio.com/docs/editor/codebasics#_multiple-selections-multicursor&#34;&gt;Multiple Cursors&lt;/a&gt; to link the images to the right timestamp on YouTube.&lt;/p&gt;
&lt;h4 id=&#34;clean-up-the-transcript&#34;&gt;Clean up the transcript&lt;/h4&gt;
&lt;p&gt;Up to here was mostly automated. This step took me an hour, though. I copied chunks of transcripts, passed it to Claude 3.5 Sonnet via Cursor with this prompt:&lt;/p&gt;
&lt;blockquote&gt;
&lt;p&gt;Clean up this webinar transcript segment. Make minimal modifications fixing spelling, grammar, punctuation, adding &amp;ldquo;quotes&amp;rdquo; where required, and combining into logical paragraphs.&lt;/p&gt;
&lt;/blockquote&gt;
&lt;p&gt;This is what gave me the bulk of the &lt;a href=&#34;https://github.com/sanand0/ai-in-education-webinar#transcript&#34;&gt;webinar transcript&lt;/a&gt;. (I&amp;rsquo;d like to automate this next.)&lt;/p&gt;
&lt;h4 id=&#34;extract-tools&#34;&gt;Extract tools&lt;/h4&gt;
&lt;p&gt;Many audience members asked for a list of tools we mentioned. So I passed ChatGPT the transcript and asked:&lt;/p&gt;
&lt;blockquote&gt;
&lt;p&gt;List all tools mentioned in this webinar&lt;/p&gt;
&lt;/blockquote&gt;
&lt;p&gt;It listed 12 tools, but I know enough to be sceptical. So&amp;hellip;&lt;/p&gt;
&lt;blockquote&gt;
&lt;p&gt;&lt;strong&gt;Anand&lt;/strong&gt;: Were any tools missed?&lt;/p&gt;
&lt;p&gt;&lt;strong&gt;ChatGPT&lt;/strong&gt;: No, the list covers all the tools mentioned in the webinar as per the transcript. If you noticed any specific tool that I missed, please let me know.&lt;/p&gt;
&lt;p&gt;&lt;strong&gt;Anand&lt;/strong&gt;: There WERE a few tools missed. Look closely. (I was bluffing, BTW.)&lt;/p&gt;
&lt;p&gt;&lt;strong&gt;ChatGPT&lt;/strong&gt;: You&amp;rsquo;re right. Upon closer review, here are the additional tools mentioned:&amp;hellip;&lt;/p&gt;
&lt;p&gt;&lt;strong&gt;Anand&lt;/strong&gt;: There are a few more that you missed.&lt;/p&gt;
&lt;p&gt;&lt;strong&gt;ChatGPT&lt;/strong&gt;: Got it. Here’s a revised list that should include all the tools mentioned:&lt;/p&gt;
&lt;/blockquote&gt;
&lt;p&gt;That generated the &lt;a href=&#34;https://github.com/sanand0/ai-in-education-webinar#tools&#34;&gt;Tools mentioned in the webinar&lt;/a&gt;.&lt;/p&gt;
&lt;h4 id=&#34;questions&#34;&gt;Questions&lt;/h4&gt;
&lt;p&gt;There were several questions in the comments. I passed them into my &lt;a href=&#34;https://colab.research.google.com/drive/19uYpWrvc1FIAYo2FVKwsLgFGntmYnd_y&#34;&gt;Topic Naming&lt;/a&gt; Colab notebook which clusters them into similar questions (I asked it to pick 40 subtopics) and then further grouped them into higher level topics, and gave names to all of these.&lt;/p&gt;
&lt;p&gt;That created the &lt;a href=&#34;https://github.com/sanand0/ai-in-education-webinar#questions&#34;&gt;list of questions people asked&lt;/a&gt;, in a categorized way.&lt;/p&gt;
&lt;h4 id=&#34;notebooklm&#34;&gt;NotebookLM&lt;/h4&gt;
&lt;p&gt;Next, I pasted the transcript into &lt;a href=&#34;https://notebooklm.google.com/notebook/44782cbd-a954-4f3b-9b32-952d68553498&#34;&gt;NotebookLM&lt;/a&gt; and repeated what our classmate &lt;a href=&#34;https://www.linkedin.com/in/raj-vadigepalli-69ba3b27/&#34;&gt;Rajanikanth&lt;/a&gt; said he did.&lt;/p&gt;
&lt;blockquote&gt;
&lt;p&gt;when I brought the transcript into NotebookLM, it suggested several questions… after clicking on those, it automatically generated answers, that I could then save into Notes. I suppose it still needs me to click on it here and there… so, I feel like I got engaged in the “learning”&lt;/p&gt;
&lt;/blockquote&gt;
&lt;p&gt;So I &amp;ldquo;clicked here and there&amp;rdquo; and generated:&lt;/p&gt;
&lt;ul&gt;
&lt;li&gt;&lt;a href=&#34;https://github.com/sanand0/ai-in-education-webinar#briefing-document&#34;&gt;A Briefing Document&lt;/a&gt;&lt;/li&gt;
&lt;li&gt;&lt;a href=&#34;https://github.com/sanand0/ai-in-education-webinar#detailed-overview&#34;&gt;A Detailed Overview&lt;/a&gt;&lt;/li&gt;
&lt;li&gt;&lt;a href=&#34;https://github.com/sanand0/ai-in-education-webinar#faq&#34;&gt;An FAQ&lt;/a&gt;&lt;/li&gt;
&lt;li&gt;&lt;a href=&#34;https://github.com/sanand0/ai-in-education-webinar#study-guide&#34;&gt;A Study Guide&lt;/a&gt;&lt;/li&gt;
&lt;/ul&gt;
&lt;p&gt;&amp;hellip; and most importantly, a &lt;a href=&#34;https://notebooklm.google.com/notebook/44782cbd-a954-4f3b-9b32-952d68553498/audio&#34;&gt;very engaging 15 minute podcast&lt;/a&gt;, which is what NotebookLM is famous for.&lt;/p&gt;
&lt;p&gt;&lt;strong&gt;Note&lt;/strong&gt;: &lt;a href=&#34;https://blog.google/technology/ai/notebooklm-update-october-2024/&#34;&gt;NotebookLM now lets you customize your podcast&lt;/a&gt;. I tried it, saying &amp;ldquo;Focus on what students and teachers can take away practically. Focus on educating rather than entertaining.&amp;rdquo; That generated a podcast that, after 5 seconds of listening, felt slightly less entertaining (duh!) so I reverted to the original.&lt;/p&gt;
&lt;h4 id=&#34;publishing&#34;&gt;Publishing&lt;/h4&gt;
&lt;p&gt;I usually publish static content as Markdown on GitHub Pages. The entire content was pushed to &lt;a href=&#34;https://github.com/sanand0/ai-in-education-webinar&#34;&gt;https://github.com/sanand0/ai-in-education-webinar&lt;/a&gt; with GitHub Pages enabled.&lt;/p&gt;
&lt;p&gt;I also created a simple &lt;a href=&#34;https://github.com/sanand0/ai-in-education-webinar&#34;&gt;index.html&lt;/a&gt; that uses &lt;a href=&#34;https://docsify.js.org/&#34;&gt;Docsify&lt;/a&gt; to convert the Markdown to HTML. I prefer this approach because it just requires adding a single HTML file to the Markdown and there is no additional deployment step. The UI is quite elegant, too.&lt;/p&gt;
&lt;h4 id=&#34;simplifying-the-workflow&#34;&gt;Simplifying the workflow&lt;/h4&gt;
&lt;p&gt;This entire workflow took me about 3 hours. Most of the manual effort went into:&lt;/p&gt;
&lt;ol&gt;
&lt;li&gt;Picking the right images (15 minutes)&lt;/li&gt;
&lt;li&gt;Cleaning up the transcript (50 minutes)&lt;/li&gt;
&lt;li&gt;Manually editing the question topics (30 minutes)&lt;/li&gt;
&lt;/ol&gt;
&lt;p&gt;If I can shorten these, I hope to transcribe and publish more of my talk videos within 15-20 minutes.&lt;/p&gt;
</description>
    </item>
    <item>
      <title></title>
      <link>https://www.s-anand.net/blog/chinese-ko-chinese-bol-sakte-hain/</link>
      <pubDate>Mon, 22 Jul 2024 03:19:27 +0000</pubDate>
      <guid>https://www.s-anand.net/blog/chinese-ko-chinese-bol-sakte-hain/</guid>
      <description>&lt;p&gt;Loved this Rocky Aur Rani Kii Prem Kahaani scene where Ranveer asks, “Chinese ko Chinese bol sakte hai?”&lt;/p&gt;
&lt;p&gt;हम बहनदी भी नहीं बोल सकते?
आंटी, मैं दिल्ली से हूँ।
मैं कैसे नहीं बहनदी बोलूं बहनदी!?
कैसा जमाना आ गया है?
फैट-ों को फैट नहीं बोल सकते, ब्लैक-ों को ब्लैक नहीं बोल सकते,
ओल्ड-ों को ओल्ड नहीं बोल सकते,
मुँह खोलने से डर लगता है मुझे!
आप मुझे बताओ, चाइनीज़ को चाइनीज़ बोल सकते हैं?&lt;/p&gt;
&lt;p&gt;Can’t we even curse “Damn it!”?
Aunty, I am from Delhi.
How can I not say, “Damn it”, damn it!?
What times are upon us?
You can’t call fat people fat, you can’t call black people black,
You can’t call old people old,
I am actually afraid to open my mouth!
Tell me, can you call Chinese people Chinese?&lt;/p&gt;
&lt;p&gt;अगर मुझे चाइनीज़ आर्डर देना है तो क्या करूँ?
फ़ोन करूँ “हाँ भाईसाहब, वह हिमाचल से राइट जो जगह है,
हाँ हाँ पड़ोस वाला मुल्क, हाँ वही कोरोना वाला, हाँ,
वहां का एक क्रिस्पी चिकन लगा दो”
अब आप बोलोगे “कोरोना वाला” नहीं बोल सकते!&lt;/p&gt;
&lt;p&gt;If I want to order Chinese, what should I do?
Should I call, “Yes bro, that place to the right of Himachal Pradesh.
Yes, yes, the neighboring country, yes, the one with Corona, yes.
Get me a crispy chicken from there”
Next, you’ll say, “You can’t say “Corona”‘!&lt;/p&gt;
&lt;p&gt;बचपन से, बचपन से दादी कहती आ रही है, “चाय पीने से काले हो जाते हैं, चाय पीने से काले हो जाते हैं”।
अब पता चला है, ये रेसिस्ट है!&lt;/p&gt;
&lt;p&gt;Since childhood, Grandma taught us, “Drinking tea turns you black. Drinking tea turns you black”.
Now I learn that’s racist!&lt;/p&gt;
&lt;p&gt;किसी ने ये भी नहीं सिखाया कि कौन सी गाली देने से मिसोजनी हो जाती है।
यह भी नहीं बताया कि गोलू को गोलू बोलने से फैट शेमिंग हो जाती है।
हमको तो यह सब नॉर्मल लगता।
हमको तो कभी रॉन्ग लगा ही नहीं।
हमको पता ही नहीं यह बोलने से सामने वाले की फीलिंग्स हर्ट हो सकती हैं!&lt;/p&gt;
&lt;p&gt;No one ever taught us which abuses become a “misogyny”.
No one ever told us that calling Chubby ‘Chubby’ is “fat-shaming”.
We all thought this was normal.
We never had an inkling it was wrong.
I didn’t know that saying this may hurt others’ feelings!&lt;/p&gt;
&lt;p&gt;Beautifully worded. Lovely acting. Nothing I could add to it.&lt;/p&gt;
&lt;p&gt;(Except: Whisper and most transcription APIs handle Hindi poorly. Claude 3.5 Sonnet does better than GPT-4o on Hindi grammar.)&lt;/p&gt;
&lt;div class=&#34;video-embed&#34;&gt;&lt;iframe src=&#34;https://www.youtube.com/embed/1h51CQwKU2k&#34; title=&#34;YouTube video&#34; loading=&#34;lazy&#34; allow=&#34;accelerometer; autoplay; clipboard-write; encrypted-media; gyroscope; picture-in-picture&#34; allowfullscreen&gt;&lt;/iframe&gt;&lt;/div&gt;
&lt;p&gt;&lt;a href=&#34;https://www.linkedin.com/feed/update/urn%3Ali%3Ashare%3A7220990805526638592&#34;&gt;LinkedIn&lt;/a&gt;&lt;/p&gt;
</description>
    </item>
  </channel>
</rss>
