<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: bobsyourbuncle</title><link>https://news.ycombinator.com/user?id=bobsyourbuncle</link><description>Hacker News RSS</description><docs>https://hnrss.org/</docs><generator>hnrss v2.1.1</generator><lastBuildDate>Wed, 22 Apr 2026 11:48:04 +0000</lastBuildDate><atom:link href="https://hnrss.org/user?id=bobsyourbuncle" rel="self" type="application/rss+xml"></atom:link><item><title><![CDATA[New comment by bobsyourbuncle in "LLMs get lost in multi-turn conversation"]]></title><description><![CDATA[
<p>Isn’t this relatively trivial to correct? Just like chain of thought reasoning replaces end tokens with “hmm” to continue the thought can’t users just replace the llm tokens whenever it starts saying “maybe they are referring to” with something like. “Let me ask a clarifying question before I proceed.”</p>
]]></description><pubDate>Thu, 15 May 2025 04:40:59 +0000</pubDate><link>https://news.ycombinator.com/item?id=43991913</link><dc:creator>bobsyourbuncle</dc:creator><comments>https://news.ycombinator.com/item?id=43991913</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=43991913</guid></item><item><title><![CDATA[New comment by bobsyourbuncle in "Stoop Coffee: A simple idea transformed my neighborhood"]]></title><description><![CDATA[
<p>How do ppl do this if they have a 9-5?</p>
]]></description><pubDate>Tue, 25 Mar 2025 22:23:05 +0000</pubDate><link>https://news.ycombinator.com/item?id=43476687</link><dc:creator>bobsyourbuncle</dc:creator><comments>https://news.ycombinator.com/item?id=43476687</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=43476687</guid></item><item><title><![CDATA[New comment by bobsyourbuncle in "How WhatsApp for business changed the world"]]></title><description><![CDATA[
<p>Where is this option? I often get errors about the video being too big and then need to crop my video length to send it. Super annoying</p>
]]></description><pubDate>Thu, 12 Dec 2024 11:35:10 +0000</pubDate><link>https://news.ycombinator.com/item?id=42398304</link><dc:creator>bobsyourbuncle</dc:creator><comments>https://news.ycombinator.com/item?id=42398304</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=42398304</guid></item><item><title><![CDATA[New comment by bobsyourbuncle in "CT Scans of New vs. Used SawStop"]]></title><description><![CDATA[
<p>Why are circular saws less prone to injury?</p>
]]></description><pubDate>Mon, 09 Dec 2024 14:37:44 +0000</pubDate><link>https://news.ycombinator.com/item?id=42366574</link><dc:creator>bobsyourbuncle</dc:creator><comments>https://news.ycombinator.com/item?id=42366574</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=42366574</guid></item><item><title><![CDATA[New comment by bobsyourbuncle in "The Curse of Recursion: Training on generated data makes models forget (2023)"]]></title><description><![CDATA[
<p>Can you clarify what you mean by using an LLM to “get a list that would even out the numbers”? If you’re doing binary classification, you need datapoints for features as well as the target class so how does an LLM synthetically create that without causing problems?</p>
]]></description><pubDate>Sun, 08 Dec 2024 02:12:56 +0000</pubDate><link>https://news.ycombinator.com/item?id=42354486</link><dc:creator>bobsyourbuncle</dc:creator><comments>https://news.ycombinator.com/item?id=42354486</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=42354486</guid></item><item><title><![CDATA[New comment by bobsyourbuncle in "Fly.io outage (restoration in progress)"]]></title><description><![CDATA[
<p>This is a good observation. Do you have any resources I can read up on to make this safer?</p>
]]></description><pubDate>Tue, 26 Nov 2024 05:25:14 +0000</pubDate><link>https://news.ycombinator.com/item?id=42242823</link><dc:creator>bobsyourbuncle</dc:creator><comments>https://news.ycombinator.com/item?id=42242823</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=42242823</guid></item><item><title><![CDATA[New comment by bobsyourbuncle in "OpenAI builds first chip with Broadcom and TSMC, scales back foundry ambition"]]></title><description><![CDATA[
<p>It’s kinda easy to check it though. Ask your friends if they use it, preferably non developer. When I find out my accounting friends and marketing friends are using it for basic analytics/promotion write ups I realized this is being used by non developers already on a regular basis. The use by non tech folks is significant for this . Crypto as a counter example of hype hasn’t lived up to its purported use case as non tech ppl I know only “own” it through an exchange.</p>
]]></description><pubDate>Wed, 30 Oct 2024 04:03:04 +0000</pubDate><link>https://news.ycombinator.com/item?id=41991862</link><dc:creator>bobsyourbuncle</dc:creator><comments>https://news.ycombinator.com/item?id=41991862</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=41991862</guid></item><item><title><![CDATA[New comment by bobsyourbuncle in "Addition Is All You Need for Energy-Efficient Language Models"]]></title><description><![CDATA[
<p>I’m new to neural nets, when should one use fp8 vs fp16 vs fp32?</p>
]]></description><pubDate>Wed, 09 Oct 2024 13:00:07 +0000</pubDate><link>https://news.ycombinator.com/item?id=41787469</link><dc:creator>bobsyourbuncle</dc:creator><comments>https://news.ycombinator.com/item?id=41787469</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=41787469</guid></item><item><title><![CDATA[New comment by bobsyourbuncle in "FTX creditors will make money on bankruptcy"]]></title><description><![CDATA[
<p>It literally says they made 900M off Anthropic in the article</p>
]]></description><pubDate>Tue, 08 Oct 2024 04:33:17 +0000</pubDate><link>https://news.ycombinator.com/item?id=41773857</link><dc:creator>bobsyourbuncle</dc:creator><comments>https://news.ycombinator.com/item?id=41773857</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=41773857</guid></item></channel></rss>