<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: marcyb5st</title><link>https://news.ycombinator.com/user?id=marcyb5st</link><description>Hacker News RSS</description><docs>https://hnrss.org/</docs><generator>hnrss v2.1.1</generator><lastBuildDate>Tue, 21 Jul 2026 20:09:12 +0000</lastBuildDate><atom:link href="https://hnrss.org/user?id=marcyb5st" rel="self" type="application/rss+xml"></atom:link><item><title><![CDATA[New comment by marcyb5st in "Kimi K3, Qwen 3.8, and Anthropic's (Potential) Unravelling"]]></title><description><![CDATA[
<p>Apparently that is what Taalas did though. Not an hardware person, so take this with a pinch of salt</p>
]]></description><pubDate>Mon, 20 Jul 2026 19:14:33 +0000</pubDate><link>https://news.ycombinator.com/item?id=48983535</link><dc:creator>marcyb5st</dc:creator><comments>https://news.ycombinator.com/item?id=48983535</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=48983535</guid></item><item><title><![CDATA[New comment by marcyb5st in "Kimi K3, Qwen 3.8, and Anthropic's (Potential) Unravelling"]]></title><description><![CDATA[
<p>No, if you burn your model into the silicon you don't need memory as the output of a layer flows through circuitry directly into the next one. No I/O to memory of any kind. You still need a bit of memory for in flight answers, but that's it.</p>
]]></description><pubDate>Mon, 20 Jul 2026 17:19:58 +0000</pubDate><link>https://news.ycombinator.com/item?id=48981810</link><dc:creator>marcyb5st</dc:creator><comments>https://news.ycombinator.com/item?id=48981810</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=48981810</guid></item><item><title><![CDATA[New comment by marcyb5st in "Kimi K3, Qwen 3.8, and Anthropic's (Potential) Unravelling"]]></title><description><![CDATA[
<p>You don't need memory because the output of the previous layer flows directly into the input gates of the next. So no I/O to/from HBM or similar<p>You still need some memory for the context, in flight answers, ... but not for the model weights and for the output of the intermediate layers. I found taalas demo here: <a href="https://chatjimmy.ai/" rel="nofollow">https://chatjimmy.ai/</a></p>
]]></description><pubDate>Mon, 20 Jul 2026 17:17:22 +0000</pubDate><link>https://news.ycombinator.com/item?id=48981763</link><dc:creator>marcyb5st</dc:creator><comments>https://news.ycombinator.com/item?id=48981763</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=48981763</guid></item><item><title><![CDATA[New comment by marcyb5st in "Migrating a production AI agent to GPT-5.6: 2.2x faster, 27% cheaper"]]></title><description><![CDATA[
<p>I believe part of the LLMOps (I don't like the term, but it is what it is) should be building a failover plan with proper testing that check tools trajectories and such. If you have these then you can sort the good enough models from cheaper to more expensive and have the failover you mentioned.<p>I saw people bulding a mapping of model->{{prompts}, {tools descriptions}, ...}, but that, to me, it feels extreme. I believe it is the model that needs to adapt to your prompts after a certain point. Models that fail to do so won't get our api requests as they will be out of the failoever roster.</p>
]]></description><pubDate>Mon, 13 Jul 2026 07:24:25 +0000</pubDate><link>https://news.ycombinator.com/item?id=48889073</link><dc:creator>marcyb5st</dc:creator><comments>https://news.ycombinator.com/item?id=48889073</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=48889073</guid></item><item><title><![CDATA[New comment by marcyb5st in "Chat Control 1.0 and 2.0 Explained"]]></title><description><![CDATA[
<p>Good thinking. Didn't think of that approach. Gonna start sending huge walls of text to DOS them then if this thing lands</p>
]]></description><pubDate>Wed, 08 Jul 2026 07:13:16 +0000</pubDate><link>https://news.ycombinator.com/item?id=48828626</link><dc:creator>marcyb5st</dc:creator><comments>https://news.ycombinator.com/item?id=48828626</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=48828626</guid></item><item><title><![CDATA[New comment by marcyb5st in "Chat Control 1.0 and 2.0 Explained"]]></title><description><![CDATA[
<p>How do they scan e2e encrypted messages? Will they force apps/OSes to have master keys/institutional backdoors to have access to the private keys?</p>
]]></description><pubDate>Wed, 08 Jul 2026 06:47:51 +0000</pubDate><link>https://news.ycombinator.com/item?id=48828445</link><dc:creator>marcyb5st</dc:creator><comments>https://news.ycombinator.com/item?id=48828445</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=48828445</guid></item><item><title><![CDATA[New comment by marcyb5st in "GLM 5.2 and the coming AI margin collapse"]]></title><description><![CDATA[
<p>It is the difference between giving an LLM an epic and say "You figure it out" and giving the single tasks' breakdown you envisioned and build incrementally on top of it.<p>With the latter you can, for example, say "Wait, this should be an interface because later on we need different concrete implementations". With the former, the agent doesn't do that, gets to the point where you actually need the flexibility interfaces give you and refactors everything to handle that. That is at least 2x the work/tokens. Multiply this for all the decision points you have to do to deliver a big piece of work and you have your bagillion tokens  consumed.</p>
]]></description><pubDate>Tue, 07 Jul 2026 13:23:42 +0000</pubDate><link>https://news.ycombinator.com/item?id=48817454</link><dc:creator>marcyb5st</dc:creator><comments>https://news.ycombinator.com/item?id=48817454</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=48817454</guid></item><item><title><![CDATA[New comment by marcyb5st in "Mystery identity of 'Green Boots' climber is finally solved after DNA test"]]></title><description><![CDATA[
<p>If you like hiking and tough, but rewarding, trails, consider the one to the base camp of Everest, or the roundtrip around the Annapurna. I did both and they are truly amazing without the craziness that has become climbing those Mountains. When I reached the basecamp it really felt like being in a weird Venice or Florence, with "tourist groups" and their guides. Crazy stuff.<p>Especially the Annapurna one, you still climb up to 5+k meters in altitude, and seeing the actual mountain still towering over you is crazy once you realize.</p>
]]></description><pubDate>Fri, 03 Jul 2026 07:27:51 +0000</pubDate><link>https://news.ycombinator.com/item?id=48771987</link><dc:creator>marcyb5st</dc:creator><comments>https://news.ycombinator.com/item?id=48771987</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=48771987</guid></item><item><title><![CDATA[New comment by marcyb5st in "US Grid Constraints: Towards 40GW+ of Behind-the-Meter Datacenter by 2028?"]]></title><description><![CDATA[
<p>High horse? Man, my country was under communist dictatorship and I didn't see a banana until I was 8 years old. However, despite that, I am aware of how lucky I got in life and I try to be mindful of those that are not as much.<p>Also, I don't believe one bit my life has been subsidized. Wealthy people don't throw money around if there is no profit to be made. At best they subsidize stuff to get you hooked before raising prices and raking the money in.<p>Additionally, I paid people to build my own house with money that I earned, and same for my car. And I now live in a country where we pay living wages and not force people to do 2-3 jobs to stay afloat. So I didn't exploit anyone as those people received months of salaries for them and their families thanks to my house.<p>The only thing I can concede to your argument is that things like my phone or laptop are cheap because people in the world are exploited while mining and in Foxconn like factories and for that I am trying to limit my tech footprint.</p>
]]></description><pubDate>Mon, 29 Jun 2026 12:45:30 +0000</pubDate><link>https://news.ycombinator.com/item?id=48718560</link><dc:creator>marcyb5st</dc:creator><comments>https://news.ycombinator.com/item?id=48718560</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=48718560</guid></item><item><title><![CDATA[New comment by marcyb5st in "US Grid Constraints: Towards 40GW+ of Behind-the-Meter Datacenter by 2028?"]]></title><description><![CDATA[
<p>Let's go back to have open air sewers while we are at it. I mean, if the goal is to bring back stuff that lowered life expectancy and reduce 50+ years of progress, I don't see why not. I really believe that as arguments go, yours is truly terrible.<p>Additionally, you proved my point that the majority really doesn't care about the folks whose quality of life worsen because someone built a datacenter close to their homes. Or the fact that turbines still pollute the air with NOx emissions (which cause lung cancer and other respiratory problems).<p>You are basically saying that it is not so bad, but (I guess) it is because you aren't close to a datacenter not depending on the grid. And I trust more hearing about live accounts of people being affected, than a "trust me bro, it is not as bad". Are the accounts cherry-picked? Maybe, but they are also supported by videos that substantiate what they are saying.</p>
]]></description><pubDate>Mon, 29 Jun 2026 11:20:57 +0000</pubDate><link>https://news.ycombinator.com/item?id=48717760</link><dc:creator>marcyb5st</dc:creator><comments>https://news.ycombinator.com/item?id=48717760</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=48717760</guid></item><item><title><![CDATA[New comment by marcyb5st in "US Grid Constraints: Towards 40GW+ of Behind-the-Meter Datacenter by 2028?"]]></title><description><![CDATA[
<p>And especially terrible for the communities that have to live with gas turbines or other local power generators as neighbors. Noise and air pollution constantly [1].<p>But fuck them, they are poor people so we don't care about them /s .<p>Additionally, people against data centers are accused of being paid by China [2]<p>[1] <a href="https://www.theguardian.com/environment/2026/feb/13/elon-musk-xai-datacenters-air-pollution-mississippi" rel="nofollow">https://www.theguardian.com/environment/2026/feb/13/elon-mus...</a>
[2] <a href="https://fortune.com/2026/06/10/kevin-oleary-trump-administration-china-data-center-sentiment-ai/" rel="nofollow">https://fortune.com/2026/06/10/kevin-oleary-trump-administra...</a></p>
]]></description><pubDate>Mon, 29 Jun 2026 09:01:12 +0000</pubDate><link>https://news.ycombinator.com/item?id=48716669</link><dc:creator>marcyb5st</dc:creator><comments>https://news.ycombinator.com/item?id=48716669</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=48716669</guid></item><item><title><![CDATA[New comment by marcyb5st in "GLM 5.2 beats Claude in our benchmarks"]]></title><description><![CDATA[
<p>Well put. I belong to the latter group as I feed small, granular tasks that I describe thoroughly to the LLM. I tried, however, to just give it a bigger scope task. Even best models produce sloppy code.<p>While the single functions/classes/structs/... can be well though out the code tends to lack cohesion, and especially maintainability. For instance, it never thinks: "I could put this logic in an interface/trait so that if the requirements change I can simply add a concrete implementation that satisfies the new requirements (and potentially use one of these for testing)".</p>
]]></description><pubDate>Mon, 29 Jun 2026 07:14:24 +0000</pubDate><link>https://news.ycombinator.com/item?id=48715866</link><dc:creator>marcyb5st</dc:creator><comments>https://news.ycombinator.com/item?id=48715866</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=48715866</guid></item><item><title><![CDATA[New comment by marcyb5st in "The deadly rise of giant trucks and SUVs"]]></title><description><![CDATA[
<p>It is not just the big cities. You go downtown even in small towns and the roads tend to be so narrow that a big SUV becomes a liability more than a vehicle. Without mentioning that I'm not even sure parking stalls fit big US cars lengthwise and they barely do widthwise</p>
]]></description><pubDate>Tue, 23 Jun 2026 21:36:55 +0000</pubDate><link>https://news.ycombinator.com/item?id=48651804</link><dc:creator>marcyb5st</dc:creator><comments>https://news.ycombinator.com/item?id=48651804</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=48651804</guid></item><item><title><![CDATA[New comment by marcyb5st in "Oracle shed about 20k roles globally in the last year"]]></title><description><![CDATA[
<p>True, but they also took huge debts to build AI DCs and not sure if the DB part of the company can cushion such a fall. According to [1] their IaaS line of business brings 4.8B USD/quarter (so say 20B/year), but they have ~120B of debt (outstanding + new debt they are trying to find people to pay for).<p>They are justifying that on commitments (500+B USD), but 300B of those are tied to OpenAI. So, if OpenAI goes belly up or at least doesn't follow their crazy growth projections, they would have to find the same amount of consumption quickly to repay the interest on said debt and eventually the principal.<p>It is a lot of money for a company the size of Oracle (~500B market cap).<p>[1] <a href="https://finance.yahoo.com/markets/stocks/articles/oracle-500-billion-backlog-vs-175401264.html" rel="nofollow">https://finance.yahoo.com/markets/stocks/articles/oracle-500...</a></p>
]]></description><pubDate>Tue, 23 Jun 2026 12:30:33 +0000</pubDate><link>https://news.ycombinator.com/item?id=48643978</link><dc:creator>marcyb5st</dc:creator><comments>https://news.ycombinator.com/item?id=48643978</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=48643978</guid></item><item><title><![CDATA[New comment by marcyb5st in "How many of the 170k English words do you know?"]]></title><description><![CDATA[
<p>For sure there is a bit of selection bias with hackernews users. Not saying we are all geniuses, but I strongly believe we are, at least, more educated than your average Joe</p>
]]></description><pubDate>Fri, 19 Jun 2026 22:23:52 +0000</pubDate><link>https://news.ycombinator.com/item?id=48603978</link><dc:creator>marcyb5st</dc:creator><comments>https://news.ycombinator.com/item?id=48603978</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=48603978</guid></item><item><title><![CDATA[New comment by marcyb5st in "How many of the 170k English words do you know?"]]></title><description><![CDATA[
<p>I think native speakers of Latin derived languages have an advantage given the proposed words in my run. The list was overly biased that way. In fact, many of the advance and grandmaster levels words are basically that. Latin derived words.<p>At least that was my experience as a native Italian speaker. My English vocabulary is good, but not great by any means and by reading books in English I know that there are plenty of words that are not derived from Latin</p>
]]></description><pubDate>Fri, 19 Jun 2026 22:17:41 +0000</pubDate><link>https://news.ycombinator.com/item?id=48603923</link><dc:creator>marcyb5st</dc:creator><comments>https://news.ycombinator.com/item?id=48603923</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=48603923</guid></item><item><title><![CDATA[New comment by marcyb5st in "Swiss parliament lifts ban on new nuclear power plants"]]></title><description><![CDATA[
<p>I think he was referring to hydro with all the mountains it is actually prime real estate for dams.</p>
]]></description><pubDate>Thu, 18 Jun 2026 15:56:37 +0000</pubDate><link>https://news.ycombinator.com/item?id=48587363</link><dc:creator>marcyb5st</dc:creator><comments>https://news.ycombinator.com/item?id=48587363</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=48587363</guid></item><item><title><![CDATA[New comment by marcyb5st in "Swiss parliament lifts ban on new nuclear power plants"]]></title><description><![CDATA[
<p>I think we (as in Switzerland) are preparing for a future in which there is not much snow melt/precipitations to fuel hydro production year round.<p>In fact, if the AMOC weakens/stops then there will be a drastic drop in precipitation across Europe and funnily enough maybe the temperature drop so much that the little snow there will be won't melt in big enough quantities.<p>Of course this is just a ban lift, meaning that there are no concrete plans to build one or more, but if there is a need to move "fast" (nuclear is not, I know) at least there is one less hurdle. I sincerily hope we invest in other technologies, especially now that Sodium batteries seem on their way to solve grid level storage, but I don't necessarily see this as a bad move per se.</p>
]]></description><pubDate>Thu, 18 Jun 2026 15:55:03 +0000</pubDate><link>https://news.ycombinator.com/item?id=48587332</link><dc:creator>marcyb5st</dc:creator><comments>https://news.ycombinator.com/item?id=48587332</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=48587332</guid></item><item><title><![CDATA[New comment by marcyb5st in "U.S. science is in chaos"]]></title><description><![CDATA[
<p>And kill the savings of what remains of the middle class. Probably they will do it though, as it is a slow thing and is not felt by the average Joe like a tax hike or loss of benefits. So the policians won't trigger an outrage by doing so.</p>
]]></description><pubDate>Wed, 17 Jun 2026 15:29:02 +0000</pubDate><link>https://news.ycombinator.com/item?id=48571852</link><dc:creator>marcyb5st</dc:creator><comments>https://news.ycombinator.com/item?id=48571852</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=48571852</guid></item><item><title><![CDATA[New comment by marcyb5st in "U.S. science is in chaos"]]></title><description><![CDATA[
<p>As an European, yeah, we probably are doing really good with basic science, but what about innovation when it comes to productivity? Why there is no AI lab (apart from Mistral) in EU? Why there is no European model (and hasn't been probably ever) in the pareto fronteer? Or any other really innovative company in the last while (I believe Spotify was the last European unicorn that transformed the landscape in the market they operate into).<p>Don't get me wrong, I rather lose the superpower race but enjoy my privacy and work benefits that folks in the US dream of. But the topic was superpower competition and I don't see the EU going anywhere in that front.<p>We are fragmented, among the top 4 EU economies 2 are struggling with debt (France & Italy), Germany economy is stagnating and the amount of bureaucracy hinders any attempt at innovation, ... .</p>
]]></description><pubDate>Wed, 17 Jun 2026 15:24:27 +0000</pubDate><link>https://news.ycombinator.com/item?id=48571784</link><dc:creator>marcyb5st</dc:creator><comments>https://news.ycombinator.com/item?id=48571784</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=48571784</guid></item></channel></rss>