<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: program_whiz</title><link>https://news.ycombinator.com/user?id=program_whiz</link><description>Hacker News RSS</description><docs>https://hnrss.org/</docs><generator>hnrss v2.1.1</generator><lastBuildDate>Wed, 18 Nov 2026 04:51:00 +0000</lastBuildDate><atom:link href="https://hnrss.org/user?id=program_whiz" rel="self" type="application/rss+xml"></atom:link><item><title><![CDATA[New comment by program_whiz in "Interpreting claims that a theorem is correct because it was checked in Lean"]]></title><description><![CDATA[
<p>Reading the friendly article:<p>It is worth noting that of the 290 OpenAI theorems which come with Lean code, around 38 of them have the property that the Lean code does not even state, let alone prove, the main theorem of the paper.<p>So even if LEAN is 100% bug-free, its still on the programmer to correctly express the logic and proofs, or they could be proving completely unrelated facts.</p>
]]></description><pubDate>Mon, 16 Nov 2026 13:15:14 +0000</pubDate><link>https://news.ycombinator.com/item?id=50517288</link><dc:creator>program_whiz</dc:creator><comments>https://news.ycombinator.com/item?id=50517288</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=50517288</guid></item><item><title><![CDATA[New comment by program_whiz in "Jonathan Blow – If AI solved coding, why is my software still the same? [video]"]]></title><description><![CDATA[
<p>The "bitter lesson" being that you need teams of dedicated scientists, engineers designing increasingly better training to squeeze more out of existing compute?  I love how people think bitter lesson was just "oh we added one more GPU and now the model is better than our hand-crafted approach".  When really its just the opposite -- models with better approaches were necessary, and models at 10% the size are showing nearly identical performance when some attention is paid to quality.  The "savory lesson" is, don't listen to the bitter lesson, it was only true when the 'hand-crafted features' experts were using were vastly over-aligned to human superstition, but if the priors are well-aligned mathematically, it really does help to have a better strategy (and more compute may not help).  The reverse-bitter-lesson every trader and investor knows: throwing more compute can often give worse results (e.g. memorization), not better.</p>
]]></description><pubDate>Sat, 14 Nov 2026 11:50:36 +0000</pubDate><link>https://news.ycombinator.com/item?id=50496265</link><dc:creator>program_whiz</dc:creator><comments>https://news.ycombinator.com/item?id=50496265</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=50496265</guid></item><item><title><![CDATA[New comment by program_whiz in "Eli Lilly CEO pushes back on Dario Amodei's medical claims"]]></title><description><![CDATA[
<p>I'm curious what disease did they find a treatment for?</p>
]]></description><pubDate>Fri, 13 Nov 2026 15:52:37 +0000</pubDate><link>https://news.ycombinator.com/item?id=50485309</link><dc:creator>program_whiz</dc:creator><comments>https://news.ycombinator.com/item?id=50485309</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=50485309</guid></item><item><title><![CDATA[New comment by program_whiz in "Doomsday: Friday, 13 November, A.D. 2026 (1960)"]]></title><description><![CDATA[
<p>yes, I read the first sentence, it was the point of contradiction in your other two.  My friend, I don't doubt you have reasons to believe what you think, but your comment as worded is self-contradicting and lacking those reasons.</p>
]]></description><pubDate>Fri, 13 Nov 2026 15:00:27 +0000</pubDate><link>https://news.ycombinator.com/item?id=50484533</link><dc:creator>program_whiz</dc:creator><comments>https://news.ycombinator.com/item?id=50484533</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=50484533</guid></item><item><title><![CDATA[New comment by program_whiz in "Eli Lilly CEO pushes back on Dario Amodei's ludicrous medical claims"]]></title><description><![CDATA[
<p>best points: drugs aren't limited by the ability to write code or do things on a computer.  Maybe if we have fully embodied robots and fully automated assembly lines to generate organic compounds and test them with extremely realistic simulations it may help slightly (there is still the FDA and production barrier).  Its just wildly unrealistic.<p>AI has made good progress in text, code, math, digital images, and digital music.  Basically troves of data that can be automated and trained on, validated, and with a purely digital interface.  Outside of that, just gathering the data will be difficult, regulated, and sparse (that is if anyone will even give it to you now).  While we can write a test to make sure we could generate drugs of the past, verifying we have generated any new working correct drugs is an RL loop with one iteration costing years and millions of dollars.</p>
]]></description><pubDate>Fri, 13 Nov 2026 14:08:41 +0000</pubDate><link>https://news.ycombinator.com/item?id=50483714</link><dc:creator>program_whiz</dc:creator><comments>https://news.ycombinator.com/item?id=50483714</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=50483714</guid></item><item><title><![CDATA[New comment by program_whiz in "Doomsday: Friday, 13 November, A.D. 2026 (1960)"]]></title><description><![CDATA[
<p>Re-read your comment it says "almost all industrialized nations are more worried about population decline than overpopulation these days. Which is why I'm not worried about superintelligence or data centers in space."<p>The only other point you made (and then violated) was "don't extrapolate too simply from past examples".   Your simple extrapolation / reasoning by analogy was "we're at peak population when projections pointed towards population explosion, therefore we won't hit superintelligence."<p>Had you used clear analysis, reasoning, or argumentation rather than just baseless extrapolation (peak population, therefore superintelligence can't happen), I wouldn't have chuckled at your self-contradicting comment.<p>"In the past, the US government has always been able to take on more debt even when people projected it would run out, therefore we'll always be able to have more intelligence until we reach super-intelligence" is the same (and equally invalid reasoning) you presented.</p>
]]></description><pubDate>Fri, 13 Nov 2026 13:59:43 +0000</pubDate><link>https://news.ycombinator.com/item?id=50483593</link><dc:creator>program_whiz</dc:creator><comments>https://news.ycombinator.com/item?id=50483593</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=50483593</guid></item><item><title><![CDATA[New comment by program_whiz in "Doomsday: Friday, 13 November, A.D. 2026 (1960)"]]></title><description><![CDATA[
<p>haha yeah, was thinking the same!  "Stop extrapolating from past data", then proceeds to extrapolate that nothing happens from a few past examples.</p>
]]></description><pubDate>Fri, 13 Nov 2026 13:07:47 +0000</pubDate><link>https://news.ycombinator.com/item?id=50482981</link><dc:creator>program_whiz</dc:creator><comments>https://news.ycombinator.com/item?id=50482981</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=50482981</guid></item><item><title><![CDATA[New comment by program_whiz in "Queens Public Library under ransomware attack"]]></title><description><![CDATA[
<p>So you'd agree it would be better to just provide those services though?  Or do you think using the library as a public restroom and shelter is ideal?</p>
]]></description><pubDate>Thu, 12 Nov 2026 14:14:08 +0000</pubDate><link>https://news.ycombinator.com/item?id=50467500</link><dc:creator>program_whiz</dc:creator><comments>https://news.ycombinator.com/item?id=50467500</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=50467500</guid></item><item><title><![CDATA[New comment by program_whiz in "Jensen Huang said to have had relationship with executive who reports to him"]]></title><description><![CDATA[
<p>And to bolster your point, people like Huang are surrounded by yes-men constantly.  When you are a nobody, people criticize and keep you in line.  Once you're running NVIDIA, people want power/money themselves, so they will tend to grovel and supplicate.  Eventually, even the humblest begin to believe their own PR.<p>Those are "external influences", but your own gravitational field has warped them. Even the president wants your money and influence, and needs your company to do well, there is literally no one who will "give it to you straight", and it warps your mind.<p>The Roman commanders would sometimes hire someone to stand behind them and remind them they were only human for this reason (to avoid natural hubris), it was seen as naturally arising in these circumstances, and needed to be combatted to remain effective.</p>
]]></description><pubDate>Thu, 12 Nov 2026 14:12:35 +0000</pubDate><link>https://news.ycombinator.com/item?id=50467457</link><dc:creator>program_whiz</dc:creator><comments>https://news.ycombinator.com/item?id=50467457</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=50467457</guid></item><item><title><![CDATA[New comment by program_whiz in "Nvidia is preparing to oust Jensen Huang"]]></title><description><![CDATA[
<p>Hanlon's razor: please don't attribute malice to us poor billionaires and the intelligence community (or hold us accountable), we're just so gosh-darn incompetent.<p>I'm tired of hearing hanlon's razor tossed around.  How about the inverse: I attribute malice if doing so makes the system better and explains your behavior.  Even if you are just a fool, your actions impact others and have the same consequences of malice.  I hold you to account like a criminal if you act like one.</p>
]]></description><pubDate>Thu, 12 Nov 2026 14:04:39 +0000</pubDate><link>https://news.ycombinator.com/item?id=50467283</link><dc:creator>program_whiz</dc:creator><comments>https://news.ycombinator.com/item?id=50467283</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=50467283</guid></item><item><title><![CDATA[New comment by program_whiz in "Writing correct software is hard and math alone won’t help us (2016)"]]></title><description><![CDATA[
<p>Formal verification is like working with stone.  If you already know the form you want, then doing so provides a rigid, long-lasting strong structure.  If you're just trying things out, then its better to sketch or work with soft clay to try the forms and see what works.<p>This is my major complaint with even less-rigid languages like rust, they kind of force you to write production-level code even if you are just brainstorming.  That's why they are terrible for games and websites, domains where correctness doesn't matter as much as feeling and iteration speed.  Sure, it would be nice if those programs had no bugs, but basically any means of increasing iteration speed and flexibility is worth it.<p>A startup or indie usually has a short runway.  Saying we're going to half that so we can make sure we don't have any memory leaks is a laugh.  It may cost 10s-100s of $K, or be the difference in launch / failure.<p>We basically would not have AI today if it weren't for loosey-goosey riddled-with-holes languages like python with all the symbolic madness they enable.  The ability of researchers to write a few lines and run it without regard to memory, processor layouts, etc. was what enabled them to write these programs at all.  Of course we've put lipstick on the pig with CUDA and torch "compiling" compute graphs, its actually such a beautiful disaster that most software runs on JS or python, two horrid languages  that have so much work to make them just performant enough.</p>
]]></description><pubDate>Thu, 12 Nov 2026 13:57:43 +0000</pubDate><link>https://news.ycombinator.com/item?id=50467151</link><dc:creator>program_whiz</dc:creator><comments>https://news.ycombinator.com/item?id=50467151</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=50467151</guid></item><item><title><![CDATA[New comment by program_whiz in "Nvidia CEO Jensen Huang's letter to staff about affair"]]></title><description><![CDATA[
<p>I'm not conflating.  I'm saying every primate species cheats, its biological.  There are species that do not (penguins), or have different mating / breeding behaviors.<p>Humans form families to care for young, but they cheat.  This is consistent across every generation and society.  We actively discourage it culturally and psychologically (as evidenced by the massive fallout Jensen is facing), so the only remaining explanation as to why it persists indefinitely is biological wiring.<p>You can google it, there are lots of references, this is a random one:
<a href="https://www.cnn.com/TECH/science/9811/24/monkeysex.enn/" rel="nofollow">https://www.cnn.com/TECH/science/9811/24/monkeysex.enn/</a></p>
]]></description><pubDate>Tue, 10 Nov 2026 14:16:06 +0000</pubDate><link>https://news.ycombinator.com/item?id=50440710</link><dc:creator>program_whiz</dc:creator><comments>https://news.ycombinator.com/item?id=50440710</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=50440710</guid></item><item><title><![CDATA[New comment by program_whiz in "What constant surveillance is doing to children and parents"]]></title><description><![CDATA[
<p>Some of that might be true, but I can assure you, people _really do_ want to prevent harm to their children.  Most of the people I know who are using this stuff do it completely voluntarily, not because of how they will be seen, but because they genuinely want to increase safety for their child.<p>Maybe its not perfect, but most parents would give everything they have if it meant saving their child's life or health, so spending a few hundred at the chance is a no-brainer.<p>Seems like "people only wear seatbelts so that they won't get tickets" type of reasoning, which is true for some, but many (myself included) like the additional safety and wear it regardless of law enforcement.  I do actually care about my safety and health, and will take reasonable measures for a chance to preserve it; and that same feeling extends to my children.</p>
]]></description><pubDate>Tue, 10 Nov 2026 13:58:34 +0000</pubDate><link>https://news.ycombinator.com/item?id=50440469</link><dc:creator>program_whiz</dc:creator><comments>https://news.ycombinator.com/item?id=50440469</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=50440469</guid></item><item><title><![CDATA[New comment by program_whiz in "Nvidia CEO Jensen Huang's letter to staff about affair"]]></title><description><![CDATA[
<p>Yes, crazy how ignoring the evidence for all of human history to support your theory means that otherwise the most rational, educated, successful people in the world are complete buffoons.<p>On the other hand if we are partially controlled by biological drives, it explains these patterns we repeatedly observe, despite the consequences.  I posit Jensen is not among "the stupidest people on earth" but that the skills that allow him to head a trillion dollar company are separate from the have-an-affair part of his biology.  Of course, you could be right, maybe Bezos, Gates, Juang, Clinton, and others are just really stupid.</p>
]]></description><pubDate>Mon, 09 Nov 2026 15:35:03 +0000</pubDate><link>https://news.ycombinator.com/item?id=50427041</link><dc:creator>program_whiz</dc:creator><comments>https://news.ycombinator.com/item?id=50427041</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=50427041</guid></item><item><title><![CDATA[New comment by program_whiz in "Nvidia CEO Jensen Huang's letter to staff about affair"]]></title><description><![CDATA[
<p>Haha, wow, yeah I'm sure the reason for all this was that Jensen couldn't find a therapist.  Aside from assuming he cheated for mental health reasons (not sure where you get that, maybe he just wanted some strange), the man could literally buy an entire therapy practice for his private use or just put one on $250k retainer to be available at all times.  I would be surprised if they didn't have coaches / mental health professionals that attend to him all the time anyway.<p>Sometimes people cheat, especially when they have lots of offers, money, free time, travel, and time away from family.  It feels good, and gives a sweet rush of dopamine, and is part of men's biological wiring and selected for over the eons.</p>
]]></description><pubDate>Mon, 09 Nov 2026 12:09:01 +0000</pubDate><link>https://news.ycombinator.com/item?id=50424500</link><dc:creator>program_whiz</dc:creator><comments>https://news.ycombinator.com/item?id=50424500</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=50424500</guid></item><item><title><![CDATA[New comment by program_whiz in "Nvidia CEO Jensen Huang's letter to staff about affair"]]></title><description><![CDATA[
<p>His PR firm definitely does, this piece gives no indication of how Jensen is actually feeling or his EQ.  I assure you he didn't draft this and push to a public website.  There are literal investigations of national security implications, as well as billions on the line over this. This is the most strategic and controlled rollout of a story you will ever read.  The wording is designed perfectly to minimize impact to stock:<p>1. take responsibility (by saying you take responsibility, hopefully there is no actual fallout from this).
2. Admit wrong doing (people hate someone who tries to cover up)
3. Talk about how much you love your family and how wonderful they are and you would never want to hurt them (relatable).
4. Write in a tone of sincere regret to avoid any public backlash (stock price depends on this).</p>
]]></description><pubDate>Mon, 09 Nov 2026 12:05:51 +0000</pubDate><link>https://news.ycombinator.com/item?id=50424471</link><dc:creator>program_whiz</dc:creator><comments>https://news.ycombinator.com/item?id=50424471</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=50424471</guid></item><item><title><![CDATA[New comment by program_whiz in "Nvidia CEO Jensen Huang's letter to staff about affair"]]></title><description><![CDATA[
<p>The GP isn't saying Jensen took the wrong action, he's saying that this is a carefully crafted message from a top PR firm, and that the posters saying how its "genuine and heartfelt, I can see he really cares" are just eating it up.  Those lines will become advertising for the agency as they land the next big client, like book reviews, but for the "publicly vulnerable executive" PR package.<p>Its not that Jensen is wrong, its a good play, the GP is just saying how sad it is to watch people eating this up like they think Jensen cracked open his laptop and drafted a note, pushed it up to the public web site, then forwarded it to axios or something.<p>There were lots of people involved in a carefully crafted rollout, the wording of which has nothing to do with how Jensen or his family are handling this. We will never know how he personally feels or how any of this really played out behind closed doors.  For all we know he and his wife are talking divorce and shouting it out and this piece is just crafted damage control.</p>
]]></description><pubDate>Mon, 09 Nov 2026 12:00:33 +0000</pubDate><link>https://news.ycombinator.com/item?id=50424414</link><dc:creator>program_whiz</dc:creator><comments>https://news.ycombinator.com/item?id=50424414</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=50424414</guid></item><item><title><![CDATA[New comment by program_whiz in "The Flash C Compiler"]]></title><description><![CDATA[
<p>A more apt analogy is heating up a frozen packaged pie, then saying it was made from scratch. Yes _you_ bought from the store, but somewhere ingredients were assembled and it was made "from scratch".</p>
]]></description><pubDate>Sat, 07 Nov 2026 12:47:58 +0000</pubDate><link>https://news.ycombinator.com/item?id=50404097</link><dc:creator>program_whiz</dc:creator><comments>https://news.ycombinator.com/item?id=50404097</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=50404097</guid></item><item><title><![CDATA[New comment by program_whiz in "Pangram Image 1.0"]]></title><description><![CDATA[
<p>In theory you are correct, in practice it works.  In theory you could use it as a training signal, in practice you cannot.  In theory, theory and practice are the same, in practice they are not.</p>
]]></description><pubDate>Fri, 06 Nov 2026 11:28:18 +0000</pubDate><link>https://news.ycombinator.com/item?id=50391048</link><dc:creator>program_whiz</dc:creator><comments>https://news.ycombinator.com/item?id=50391048</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=50391048</guid></item><item><title><![CDATA[New comment by program_whiz in "Pangram Image 1.0"]]></title><description><![CDATA[
<p>Also you aren't getting solid gradients, it would be a simple yes/no (or confidence score), as you don't have their model.  GANs failed to generalize, even with full access to the discriminator model (providing robust gradients to all weights), its extremely difficult to train in what is basically a reinforcement learning loop over a possibly extremely spiky, nonlinear terrain.<p>Think of the resources it takes to train good text generators, with perfect knowledge of the next token to generate.  Now imagine to trying to do it with a stochastic output of yes/no only (needing to guess proper outputs, and the signal you're getting isn't even deterministic / consistent).  You don't even have the distribution over output tokens, you only have yes/no based on the tokens you did generate (a far weaker signal).</p>
]]></description><pubDate>Fri, 06 Nov 2026 11:27:08 +0000</pubDate><link>https://news.ycombinator.com/item?id=50391040</link><dc:creator>program_whiz</dc:creator><comments>https://news.ycombinator.com/item?id=50391040</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=50391040</guid></item></channel></rss>