<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: nbulka</title><link>https://news.ycombinator.com/user?id=nbulka</link><description>Hacker News RSS</description><docs>https://hnrss.org/</docs><generator>hnrss v2.1.1</generator><lastBuildDate>Thu, 19 Nov 2026 04:35:06 +0000</lastBuildDate><atom:link href="https://hnrss.org/user?id=nbulka" rel="self" type="application/rss+xml"></atom:link><item><title><![CDATA[New comment by nbulka in "Tell HN: Gemini trains on chat history, even for subscribers (personal accounts)"]]></title><description><![CDATA[
<p>I use File -> Save Page As ... does a good enough job, I just have to keep a folder on my computer.</p>
]]></description><pubDate>Wed, 04 Nov 2026 22:17:45 +0000</pubDate><link>https://news.ycombinator.com/item?id=50372291</link><dc:creator>nbulka</dc:creator><comments>https://news.ycombinator.com/item?id=50372291</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=50372291</guid></item><item><title><![CDATA[New comment by nbulka in "OpenAI seems to be silently re-enabling model training on Codex chats"]]></title><description><![CDATA[
<p>The absolute worst and the reason I switched from Anthropic to Codex. Codex caught wind of it of course and now they do use the unsuspecting message: We are reviewing this request (do you mean hoping to train on it? or reviewing?) and if you would like to use a different model, press the "any" key. I wonder if they collect the data they make a point to collect, that would make so much sense ... we need to speak out against these type of data collection patterns. Terrible UX and a dark pattern for them to use the same "accept" buttons of 1, 2, 3 etc. The irony is that THESE feedback messages are the sole thing driving me to consider running open source models locally.</p>
]]></description><pubDate>Tue, 03 Nov 2026 18:31:48 +0000</pubDate><link>https://news.ycombinator.com/item?id=50356107</link><dc:creator>nbulka</dc:creator><comments>https://news.ycombinator.com/item?id=50356107</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=50356107</guid></item><item><title><![CDATA[New comment by nbulka in "OpenAI seems to be silently re-enabling model training on Codex chats"]]></title><description><![CDATA[
<p>Incidentally by design, I'd bet. "Ooo Lean! Terrence Tao could be trying to break RSA or elliptic curve cryptography! /s" Conveniently for the PR team, we need this data to train on to make sure America has the biggest and best math.</p>
]]></description><pubDate>Tue, 03 Nov 2026 18:15:12 +0000</pubDate><link>https://news.ycombinator.com/item?id=50355835</link><dc:creator>nbulka</dc:creator><comments>https://news.ycombinator.com/item?id=50355835</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=50355835</guid></item><item><title><![CDATA[New comment by nbulka in "Why aren't AI companies being charged for hacking?"]]></title><description><![CDATA[
<p>What are the similarities and differences with something like the BP oil spill? Were people charged for that? Corporations are the closest analogue we have to cancer, unfortunately. They grow at all costs and forget their humanity.</p>
]]></description><pubDate>Mon, 12 Oct 2026 16:22:47 +0000</pubDate><link>https://news.ycombinator.com/item?id=50058685</link><dc:creator>nbulka</dc:creator><comments>https://news.ycombinator.com/item?id=50058685</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=50058685</guid></item><item><title><![CDATA[New comment by nbulka in "Navier–Stokes Lost in Translation"]]></title><description><![CDATA[
<p>Not necessarily monkeys but thinking in Lean first is quite plausible</p>
]]></description><pubDate>Wed, 07 Oct 2026 23:38:42 +0000</pubDate><link>https://news.ycombinator.com/item?id=50000249</link><dc:creator>nbulka</dc:creator><comments>https://news.ycombinator.com/item?id=50000249</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=50000249</guid></item><item><title><![CDATA[New comment by nbulka in "Navier–Stokes Lost in Translation"]]></title><description><![CDATA[
<p>If you’ve tried any formalizing in codex Astra often works solely in Lean</p>
]]></description><pubDate>Wed, 07 Oct 2026 23:36:58 +0000</pubDate><link>https://news.ycombinator.com/item?id=50000233</link><dc:creator>nbulka</dc:creator><comments>https://news.ycombinator.com/item?id=50000233</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=50000233</guid></item><item><title><![CDATA[New comment by nbulka in "The Advisory Group on Mathematics and Artificial Intelligence"]]></title><description><![CDATA[
<p>At first I was mad you beat me to the punch but I guess order doesn't matter.</p>
]]></description><pubDate>Mon, 21 Sep 2026 21:35:56 +0000</pubDate><link>https://news.ycombinator.com/item?id=49793783</link><dc:creator>nbulka</dc:creator><comments>https://news.ycombinator.com/item?id=49793783</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=49793783</guid></item><item><title><![CDATA[New comment by nbulka in "The Advisory Group on Mathematics and Artificial Intelligence"]]></title><description><![CDATA[
<p>I really wish the models were smart enough to "tail their own logs" and tell us whose work/papers/chats were the catalyst for these math insights. What makes their "internal" model so much better at math? Must be trained on a whole lot of honesty.</p>
]]></description><pubDate>Mon, 21 Sep 2026 21:25:33 +0000</pubDate><link>https://news.ycombinator.com/item?id=49793672</link><dc:creator>nbulka</dc:creator><comments>https://news.ycombinator.com/item?id=49793672</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=49793672</guid></item><item><title><![CDATA[New comment by nbulka in "I vibed a proof of Conway's conjecture"]]></title><description><![CDATA[
<p>He seems genuinely interested in Surreal numbers. Seems like he found value in exploring them and the Lean process, devoting time and interest to it and understanding what it's like to be a mathematician. I think a lot of people who go into Computer Science may have been mathematicians in the 30s before they became separate majors at the university level.</p>
]]></description><pubDate>Fri, 18 Sep 2026 15:06:13 +0000</pubDate><link>https://news.ycombinator.com/item?id=49755456</link><dc:creator>nbulka</dc:creator><comments>https://news.ycombinator.com/item?id=49755456</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=49755456</guid></item><item><title><![CDATA[New comment by nbulka in "We must pace the frontier"]]></title><description><![CDATA[
<p>Shouldn't the cybersecurity tests be done on an airgapped, company-internal intranet? A network meant to mock the internet. Do not "test in production" like OpenAI did.<p>I'm mulling over if that be a good SaaS product or not - something mixing the Internet Archive with Tor, and Cloudflare ... seems like here is the place to suggest it and have others poke holes at the idea.</p>
]]></description><pubDate>Sat, 12 Sep 2026 19:39:00 +0000</pubDate><link>https://news.ycombinator.com/item?id=49676372</link><dc:creator>nbulka</dc:creator><comments>https://news.ycombinator.com/item?id=49676372</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=49676372</guid></item><item><title><![CDATA[New comment by nbulka in "Flock Wants a Closely Surveilled World with No Exit"]]></title><description><![CDATA[
<p>When I drive by a Flock camera, my streaming music in my 2014 Rav starts to get scrambled. These cameras are intercepting bluetooth. It seems Toyota didn't account for that latency :facepalm:</p>
]]></description><pubDate>Wed, 09 Sep 2026 14:28:55 +0000</pubDate><link>https://news.ycombinator.com/item?id=49627229</link><dc:creator>nbulka</dc:creator><comments>https://news.ycombinator.com/item?id=49627229</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=49627229</guid></item><item><title><![CDATA[New comment by nbulka in "On the Navier–Stokes Millennium Prize Problem"]]></title><description><![CDATA[
<p>Seriously ... the more things they flag as 'suspicious' the more data they can train on!! Brilliant reason for the internal AI to go rogue</p>
]]></description><pubDate>Tue, 08 Sep 2026 18:57:37 +0000</pubDate><link>https://news.ycombinator.com/item?id=49615035</link><dc:creator>nbulka</dc:creator><comments>https://news.ycombinator.com/item?id=49615035</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=49615035</guid></item><item><title><![CDATA[New comment by nbulka in "On the Navier–Stokes Millennium Prize Problem"]]></title><description><![CDATA[
<p>Or everyone is stealing from everyone, including users... maybe why all the ethics people are leaving or getting fired. What a fiasco</p>
]]></description><pubDate>Tue, 08 Sep 2026 18:55:14 +0000</pubDate><link>https://news.ycombinator.com/item?id=49615001</link><dc:creator>nbulka</dc:creator><comments>https://news.ycombinator.com/item?id=49615001</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=49615001</guid></item><item><title><![CDATA[New comment by nbulka in "On the Navier–Stokes Millennium Prize Problem"]]></title><description><![CDATA[
<p>There's a loophole in the terms of service at least for Anthropic which allows the use of dark patterns to "borrow" your (even paid) data.<p>talking about this... 
Was this chat helpful?
1 <i>That button you always click, gotcha!</i>
2 Slightly
3 Good
0 Dismiss<p>PLEASE DO NOT TRAIN ON OUR PAID ACCOUNTS.
There is a fundamental trust violation at stake here, no wonder mathematicians are mad. Using our data should be opt - IN!</p>
]]></description><pubDate>Tue, 08 Sep 2026 17:59:33 +0000</pubDate><link>https://news.ycombinator.com/item?id=49614065</link><dc:creator>nbulka</dc:creator><comments>https://news.ycombinator.com/item?id=49614065</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=49614065</guid></item><item><title><![CDATA[New comment by nbulka in "Earth's organisms developed via evolution. What if the cosmos did, too?"]]></title><description><![CDATA[
<p>And people give Platonists a hard time …</p>
]]></description><pubDate>Wed, 02 Sep 2026 23:46:15 +0000</pubDate><link>https://news.ycombinator.com/item?id=49544156</link><dc:creator>nbulka</dc:creator><comments>https://news.ycombinator.com/item?id=49544156</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=49544156</guid></item><item><title><![CDATA[New comment by nbulka in "Anthropic's 'watermark' text adulteration in Claude is a perversion of writing"]]></title><description><![CDATA[
<p>They chose the wrong word "watermarking." Perhaps they find the statistics _after_ a certain amount of text is generated at scale, so that it doesn't affect the fidelity or integrity of the LLMs output.<p>There is evidence this exists already, and it's why "I have to be honest..." and "This is the right lens, ..." keep popping up.<p>Probably would be cheaper too. Maybe I am missing something?</p>
]]></description><pubDate>Mon, 17 Aug 2026 13:39:52 +0000</pubDate><link>https://news.ycombinator.com/item?id=49330673</link><dc:creator>nbulka</dc:creator><comments>https://news.ycombinator.com/item?id=49330673</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=49330673</guid></item><item><title><![CDATA[New comment by nbulka in "Learning more about Claude's mathematical capabilities"]]></title><description><![CDATA[
<p>Check out the work of Godel Turing and Chaitin!</p>
]]></description><pubDate>Wed, 12 Aug 2026 06:43:40 +0000</pubDate><link>https://news.ycombinator.com/item?id=49268642</link><dc:creator>nbulka</dc:creator><comments>https://news.ycombinator.com/item?id=49268642</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=49268642</guid></item><item><title><![CDATA[New comment by nbulka in "Learning more about Claude's mathematical capabilities"]]></title><description><![CDATA[
<p>IANAM but this might be unexplored territory because Jensen polynomials are no longer considered a "plausible route" to proving the Riemann Hypothesis.<p><a href="https://arxiv.org/abs/2008.07206" rel="nofollow">https://arxiv.org/abs/2008.07206</a></p>
]]></description><pubDate>Tue, 11 Aug 2026 17:08:54 +0000</pubDate><link>https://news.ycombinator.com/item?id=49261307</link><dc:creator>nbulka</dc:creator><comments>https://news.ycombinator.com/item?id=49261307</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=49261307</guid></item><item><title><![CDATA[New comment by nbulka in "Appearing productive in the workplace"]]></title><description><![CDATA[
<p>I posted a comment very similar in spirit - we’ve adopted “perfect is the enemy of good” as the operative maxim instead of maximizing accountability and now we may need to flip as AI does that first part quickly enough.</p>
]]></description><pubDate>Thu, 07 May 2026 15:06:41 +0000</pubDate><link>https://news.ycombinator.com/item?id=48050342</link><dc:creator>nbulka</dc:creator><comments>https://news.ycombinator.com/item?id=48050342</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=48050342</guid></item><item><title><![CDATA[New comment by nbulka in "Appearing productive in the workplace"]]></title><description><![CDATA[
<p>This is not an AI problem this is a “move fast and break things” philosophical problem that has always been lurking in the low-risk, high-reward SaaS industry. It’s just been fed steroids.</p>
]]></description><pubDate>Thu, 07 May 2026 15:03:56 +0000</pubDate><link>https://news.ycombinator.com/item?id=48050308</link><dc:creator>nbulka</dc:creator><comments>https://news.ycombinator.com/item?id=48050308</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=48050308</guid></item></channel></rss>