<rss version="2.0" xmlns:dc="http://purl.org/dc/elements/1.1/" xmlns:atom="http://www.w3.org/2005/Atom"><channel><title>Hacker News: quickthoughts</title><link>https://news.ycombinator.com/user?id=quickthoughts</link><description>Hacker News RSS</description><docs>https://hnrss.org/</docs><generator>hnrss v2.1.1</generator><lastBuildDate>Sun, 14 Jun 2026 22:28:01 +0000</lastBuildDate><atom:link href="https://hnrss.org/user?id=quickthoughts" rel="self" type="application/rss+xml"></atom:link><item><title><![CDATA[New comment by quickthoughts in "AI Coding at Home Without Going Broke"]]></title><description><![CDATA[
<p>Ha just wrote a post[1] about a sort of 4th option - max out cheap compute to create more tangible things that can be used/run locally.<p>1: <a href="https://news.ycombinator.com/item?id=48519181">https://news.ycombinator.com/item?id=48519181</a></p>
]]></description><pubDate>Sat, 13 Jun 2026 17:13:11 +0000</pubDate><link>https://news.ycombinator.com/item?id=48519257</link><dc:creator>quickthoughts</dc:creator><comments>https://news.ycombinator.com/item?id=48519257</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=48519257</guid></item><item><title><![CDATA[New comment by quickthoughts in "Vercel Drop"]]></title><description><![CDATA[
<p>what does this actually do? claude output -> drop -> app that you can't really change?</p>
]]></description><pubDate>Sat, 13 Jun 2026 17:08:19 +0000</pubDate><link>https://news.ycombinator.com/item?id=48519205</link><dc:creator>quickthoughts</dc:creator><comments>https://news.ycombinator.com/item?id=48519205</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=48519205</guid></item><item><title><![CDATA[A case for AI-pragmatism – leverage cheap compute while it's here]]></title><description><![CDATA[
<p>Article URL: <a href="https://quickthoughts.ca/posts/a-case-for-pragmatism-in-ai/">https://quickthoughts.ca/posts/a-case-for-pragmatism-in-ai/</a></p>
<p>Comments URL: <a href="https://news.ycombinator.com/item?id=48519181">https://news.ycombinator.com/item?id=48519181</a></p>
<p>Points: 3</p>
<p># Comments: 0</p>
]]></description><pubDate>Sat, 13 Jun 2026 17:06:24 +0000</pubDate><link>https://quickthoughts.ca/posts/a-case-for-pragmatism-in-ai/</link><dc:creator>quickthoughts</dc:creator><comments>https://news.ycombinator.com/item?id=48519181</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=48519181</guid></item><item><title><![CDATA[New comment by quickthoughts in "Ask HN: Are most corporate SWE jobs performative?"]]></title><description><![CDATA[
<p>Are most corporations preformative? (yes)</p>
]]></description><pubDate>Wed, 10 Jun 2026 19:26:04 +0000</pubDate><link>https://news.ycombinator.com/item?id=48481363</link><dc:creator>quickthoughts</dc:creator><comments>https://news.ycombinator.com/item?id=48481363</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=48481363</guid></item><item><title><![CDATA[New comment by quickthoughts in "Gemma 4 12B: A unified, encoder-free multimodal model"]]></title><description><![CDATA[
<p>I use small models like Gemma to improve transcriptions from ASR models amongst other micro-tasks. I actually built out a fine-tuning whisper pipeline with all local (smaller) models meaning no cloud/big-tech co is able to train/sell my (private) data.<p>Repo is <a href="https://github.com/Rebreda/listenr" rel="nofollow">https://github.com/Rebreda/listenr</a> - mainly geared toward Whisper fine-tuning, AMD hardware and local inference</p>
]]></description><pubDate>Wed, 03 Jun 2026 18:32:22 +0000</pubDate><link>https://news.ycombinator.com/item?id=48387870</link><dc:creator>quickthoughts</dc:creator><comments>https://news.ycombinator.com/item?id=48387870</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=48387870</guid></item><item><title><![CDATA[New comment by quickthoughts in "Apple's privacy invading tech is likely going mainstream"]]></title><description><![CDATA[
<p>I don't think this will be the case, image and video processing typically is more computation hungry, especially in realtime contexts. Apple already pegged Google as their cloud-provider of choice when it comes to LLMs/AI compute.<p><a href="https://www.business-standard.com/technology/tech-news/apple-says-gemini-powered-siri-will-run-on-its-private-cloud-what-it-means-126013000427_1.html" rel="nofollow">https://www.business-standard.com/technology/tech-news/apple...</a></p>
]]></description><pubDate>Mon, 11 May 2026 11:45:00 +0000</pubDate><link>https://news.ycombinator.com/item?id=48093759</link><dc:creator>quickthoughts</dc:creator><comments>https://news.ycombinator.com/item?id=48093759</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=48093759</guid></item><item><title><![CDATA[Apple's privacy invading tech is likely going mainstream]]></title><description><![CDATA[
<p>Article URL: <a href="https://quickthoughts.ca/posts/self-owning-ai-tech/">https://quickthoughts.ca/posts/self-owning-ai-tech/</a></p>
<p>Comments URL: <a href="https://news.ycombinator.com/item?id=48090317">https://news.ycombinator.com/item?id=48090317</a></p>
<p>Points: 6</p>
<p># Comments: 6</p>
]]></description><pubDate>Mon, 11 May 2026 02:12:31 +0000</pubDate><link>https://quickthoughts.ca/posts/self-owning-ai-tech/</link><dc:creator>quickthoughts</dc:creator><comments>https://news.ycombinator.com/item?id=48090317</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=48090317</guid></item><item><title><![CDATA[New comment by quickthoughts in "FBI is buying location data to track US citizens, director confirms"]]></title><description><![CDATA[
<p>It's everyone. Especially google, but all the big tech companies play in the same pool. Amazon, Google, Apple, Meta etc make money selling ads, which ultimate enables the tools that result data harvesting from everyone across the internet. I wrote a little data investigation [1] (mostly finished) that show cases how every major news organization across the globe I scanned had some level of data collection integrated. This is just one industry, but its important (as it connects back to the incentives these media organizations have, which is to make money by selling ads at any cost). The eff also released an angle in how the bidding process to buy ads is itself a massive privacy nightmare[2]<p>[1] <a href="https://quickthoughts.ca/autotracko/" rel="nofollow">https://quickthoughts.ca/autotracko/</a>
[2] <a href="https://www.eff.org/deeplinks/2026/03/targeted-advertising-gives-your-location-government-just-ask-cbp" rel="nofollow">https://www.eff.org/deeplinks/2026/03/targeted-advertising-g...</a></p>
]]></description><pubDate>Thu, 19 Mar 2026 00:58:22 +0000</pubDate><link>https://news.ycombinator.com/item?id=47433428</link><dc:creator>quickthoughts</dc:creator><comments>https://news.ycombinator.com/item?id=47433428</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=47433428</guid></item><item><title><![CDATA[New comment by quickthoughts in "My Journey to a reliable and enjoyable locally hosted voice assistant (2025)"]]></title><description><![CDATA[
<p>I've been working on the flip side of this with ASR models, but the problem space is the same, conversational/real-world data is needed. Whisper often mistook actual words I say and hallucinate all the time when speaking technical jargon. The solution is to fine-tuning whisper with my own data. Hardest part imo was getting the actual data, which in turn got me to build listenr (<a href="https://github.com/rebreda/listenr" rel="nofollow">https://github.com/rebreda/listenr</a>).It's an always-on VAD-based audio dataset builder. Could be used for building conversational/real-world voice datasets for TTS models too?<p>After getting it working i was get motivation to actually able to build out the full fine-tuning pipeline. I wrote a little post about it all <a href="https://quickthoughts.ca/posts/listenr-asr-training-data-problem/" rel="nofollow">https://quickthoughts.ca/posts/listenr-asr-training-data-pro...</a></p>
]]></description><pubDate>Mon, 16 Mar 2026 21:32:06 +0000</pubDate><link>https://news.ycombinator.com/item?id=47405202</link><dc:creator>quickthoughts</dc:creator><comments>https://news.ycombinator.com/item?id=47405202</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=47405202</guid></item><item><title><![CDATA[New comment by quickthoughts in "Coffee for people who don't like coffee"]]></title><description><![CDATA[
<p>Plus one for the IKEA pour over. Its cheap and functional, perfect for travel too.</p>
]]></description><pubDate>Wed, 14 May 2025 04:22:08 +0000</pubDate><link>https://news.ycombinator.com/item?id=43980777</link><dc:creator>quickthoughts</dc:creator><comments>https://news.ycombinator.com/item?id=43980777</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=43980777</guid></item><item><title><![CDATA[New comment by quickthoughts in "Ask HN: Cursor or Windsurf?"]]></title><description><![CDATA[
<p>VSCode with auto complete and Gemini 2.5 Pro in a standalone chat (pick any interface that works for you, eg librechat, vertex etc). The agents-in-a-IDE experience is hella slow in my opinion.<p>Plus its less about the actual code generation and more about how to use it effectively. I wrote a simple piece on how I use it to automate the boring parts of dev work to great effect <a href="https://quickthoughts.ca/posts/automate-smarter-maximizing-reliable-workflows-with-llm-generated-scripts" rel="nofollow">https://quickthoughts.ca/posts/automate-smarter-maximizing-r...</a>)</p>
]]></description><pubDate>Mon, 12 May 2025 11:33:44 +0000</pubDate><link>https://news.ycombinator.com/item?id=43961871</link><dc:creator>quickthoughts</dc:creator><comments>https://news.ycombinator.com/item?id=43961871</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=43961871</guid></item></channel></rss>