<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: mlikelihood</title><link>https://news.ycombinator.com/user?id=mlikelihood</link><description>Hacker News RSS</description><docs>https://hnrss.org/</docs><generator>hnrss v2.1.1</generator><lastBuildDate>Thu, 19 Nov 2026 05:09:33 +0000</lastBuildDate><atom:link href="https://hnrss.org/user?id=mlikelihood" rel="self" type="application/rss+xml"></atom:link><item><title><![CDATA[New comment by mlikelihood in "Harvard researchers find AI coding agents shifted the bottleneck to code review"]]></title><description><![CDATA[
<p>So many people are still in denial phase. Why still try to review?<p>We've built auto-reviewers (I'm sure you can do it on any grok bot / dot / solutions etc.) and auto-review resolvers -- always-on agents running in sandbox and review and RESOLVE comments (with new commits) on any author's github profile. You can tune the prompts for nit-picking, pushing back, etc.<p>Then you can do regular audit to see how many false positive and negatives are there in all the PRs merged.<p>Over time, we've noticed less and less P0 bugs are even caught, and the for those P1 and P2 issues, auto-resolving is perfectly fine/low to no risk at all if it's just another "polishment" like edge case handling that's not so common.<p>Then the problem now shifts to architecture design, task decomposition, etc.</p>
]]></description><pubDate>Wed, 18 Nov 2026 19:32:51 +0000</pubDate><link>https://news.ycombinator.com/item?id=50550954</link><dc:creator>mlikelihood</dc:creator><comments>https://news.ycombinator.com/item?id=50550954</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=50550954</guid></item><item><title><![CDATA[New comment by mlikelihood in "Ask HN: What's your current AI agentic setup?"]]></title><description><![CDATA[
<p>this. Fable is an overkill in most cases given its high price. It's better, just not as better for its cost.<p>Have been using Codex + Claude (their $100 plans) and for some time thought Claude was too slow and wordy so using Sol for day to day; then Opus 5.5 turns out to be still the best in quality and they've caught up in speed recently as well.<p>I think it's always best to have multiple agents review or discuss on the problem then delegate to a faster/lower cost model for execution.</p>
]]></description><pubDate>Tue, 10 Nov 2026 22:13:20 +0000</pubDate><link>https://news.ycombinator.com/item?id=50447027</link><dc:creator>mlikelihood</dc:creator><comments>https://news.ycombinator.com/item?id=50447027</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=50447027</guid></item></channel></rss>