Hacker News: eitally

New comment by eitally in "A third of Perplexity's citations don't contain the number they're cited for"

eitally — Wed, 02 Sep 2026 17:25:03 +0000

This has become a common pattern in my Claude & Gemini usage. Always require citations, then check those citations to validate they actually contain the information/data the LLM's output claims. Claude, in particular, seems to make massive logic jumps and trust tertiary data sources way more than it should.

New comment by eitally in "AWS Acquires DuckLabs"

eitally — Wed, 26 Aug 2026 15:12:46 +0000

Honestly, there's just a huge amount of thrash across the cloud & SaaS industries as a whole. It's not a lot better at Google (way behind OAI & Anthropic on coding & local AI harnesses), Microsoft (no meaningful frontier model R&D to speak of, and a constantly eroding Windows business), the neoclouds (impossible to source adequate infrastructure and are existentially threatened by the hyperscalers (especially Google) if they can't keep up with demand, chipmakers (NVIDIA vs AMD vs Qualcomm vs Apple vs the long tail of specialty shops like Cerebras), data center pure plays (legislative & community pushback, natural resources, infrastructure availability), and the SaaS folks ("let's see if AI can do it").

It's both a great time to be gold mining in tech, and also a terrible time to be a bit employee.

New comment by eitally in "Cerebras CS-4"

eitally — Wed, 19 Aug 2026 03:35:01 +0000

Maybe, but GPU is just one aspect of NVIDIA's dominance. If you are buying Vera Rubin GPUs, you're getting an NVL72 rack, which is only one of several racks that you're probably buying. You'll also need your NVIDIA racks with NVIDIA networking & storage gear, too. At the end of the day, they're "vertically integrated" for your accelerated computing data center (e.g. the "AI Factory"). This doesn't even count the software layer, where CUDA + CUDA-X (not to mention the software for all the sysadmin pieces) has a huge first mover advantage over anyone else.

New comment by eitally in "How do I permanently disable random Google Photos popup to backup photos? (2024)"

eitally — Mon, 17 Aug 2026 13:43:40 +0000

How did you effectively download 500GB of photos? Did you just use Takeout?

New comment by eitally in "Seven books I keep close because I love them"

eitally — Sat, 15 Aug 2026 02:24:51 +0000

I think it's arguable that this "intuition" you call knowledge could very easily also be called "opinion".

New comment by eitally in "Muse Glimmer: 30B-parameter model optimized for always-on local agent workflows"

eitally — Mon, 10 Aug 2026 16:37:21 +0000

The interesting thing is that China began as an agrarian economy and quickly modernized via influence/direction of the CCP.

The United States used to have a dominant middle class that was geographically distributed (cities & rural areas inclusive), but the advent of the tech economy has also been having a similar effect here as it did in China: massive wealth accumulation in Tier 1-3 cities and everyone else being largely left behind.

It'll be interesting to see if there's convergence in the next decade or so, especially with the GOP reducing regulation and increasing the explicit capitalist priorities here. TBH, though, as a generally well-informed political outsider, the feeling I have is that the US government is slow, bulky, inefficient, balkanized and overall poorly run compared to the Chinese government. We'll likely either slowly improve authoritarian efficiency and become more like the CCP ... or we'll pivot left and move more toward the EU model, but we're floundering around right now paying lip service to both.

New comment by eitally in "Managing AI Coding Costs at Scale"

eitally — Fri, 07 Aug 2026 22:49:02 +0000

As a business user, the same thing is true for non-code documents. The biggest exertion is reducing the excessive slop down to concise, clear points.

New comment by eitally in "USA Today Co., partners with Palantir to analyze audience data"

eitally — Fri, 07 Aug 2026 16:54:39 +0000

My question is this: why is Palantir required for this kind of fairly basic data analysis? Couldn't they use much cheaper, probably open platforms for data science & analytics?

New comment by eitally in "Not hiring junior engineers won't solve the problem you think you have"

eitally — Wed, 05 Aug 2026 16:05:50 +0000

FWIW, your pessimism isn't entirely well-founded. If you're in a hot, tech-centric market, then yes, staff will be far more inclined to job hop for comp increases. But if you're hiring remote or you're located outside one of the very few tech hubs, employees value quality of life and workplace culture/relationships much more highly relative to comp. E.g. If you're hiring a SWE in Nashville or Portland or Columbus they're going to behave differently than if you're hiring in the Bay Area, Seattle area, NYC, Boston or Austin.

New comment by eitally in "The American Grilled Cheese Sandwich Essay (2024)"

eitally — Fri, 31 Jul 2026 04:38:06 +0000

I disagree, or rather I don't care if I'm wrong but there are likely millions on my side. I'm 49 years old, grew up eating grilled cheese (American cheese on Pepperidge Farms white bread), often with sliced onion or tomato, and mustard. I had never even heard of a melt until I was probably 35, and even so, I still call a doctored up grilled cheese sandwich a grilled cheese sandwich. It ceases to be grilled cheese only if meat is added.

New comment by eitally in "Terence Tao's ChatGPT conversation about the Jacobian Conjecture counterexample"

eitally — Thu, 23 Jul 2026 19:22:03 +0000

I am not impressed with that steering of the AI. In the case of Terry Tao, he clearly held a solid understanding of the problem domain and could also understand the outputs of the LLM at an expert level. He asked astute, topical questions that specifically probed areas of the problem that were either unclear or insufficiently specified, and ultimately ended up at a place where (presumably) he both understood the Jacobian Conjecture Counterexample exhaustively, and how it was derived. My impression is that through his learnings during the chat, he became sufficiently well-educated in this subdomain to apply some of these lessons to other mathematical pursuits.

In your counterexample, the human just repeatedly asked the model to "try harder". Not remotely the same.

New comment by eitally in "Gemini 3.6 Flash, 3.5 Flash-Lite, and 3.5 Flash Cyber"

eitally — Wed, 22 Jul 2026 03:41:39 +0000

Google already owns 14% of Anthropic.

New comment by eitally in "The state of open source AI"

eitally — Fri, 17 Jul 2026 20:25:56 +0000

I think this is possibly true, but the other piece of the equation is tooling. Right now, Anthropic has by far the best tooling around (for both SWEs and non-technical users), and a huge ecosystem of integrated ISVs. I'm not suggesting it will happen, but if Anthropic decided to provide options of using models beyond Claude, they'd still have a significant moat.

New comment by eitally in "Nvidia, CoreWeave, and Nebius: Inside the Circular Financing of the GPU Boom"

eitally — Sat, 11 Jul 2026 23:44:28 +0000

NVIDIA's $2b into CoreWeave was a stock purchase.

https://investors.coreweave.com/news/news-details/2026/NVIDI...

New comment by eitally in "Apple Silicon Exec Explains Mac Mini AI Demand and On-Device Future"

eitally — Fri, 10 Jul 2026 14:22:59 +0000

I think so, too, and I think it'll end up being a race between Apple & NVIDIA (or NVIDIA partners) to see who realizes this first. It would probably be easier for Apple to do it because it wouldn't require a form factor adjustment [over the Mac Studio they already have]. That said, NVIDIA already offers chipsets for both the lower end (DGX Spark with Vera + GB10, at roughly the $4500 price point) and higher end (DGX Station with Vera + GB300, for $85-100k). The DGX Station is equivalent to ~5-6 RTX6000 GPUs attached to a mid-range CPU server, but far more than most individual developers would want or need. I've heard through the grapevine that NVIDIA's received consistent feedback that they need something like a "GB20" that slots above the Spark/GB10 and can simultaneously run larger models for inference while hosting a dev environment on the same box. You can daisy-chain Sparks just like you can daisy-chain Mac Minis, but you're still constrained on model performance based on what a single device can accommodate.

New comment by eitally in "Apple to increase spend with Broadcom to produce billions more U.S. chips"

eitally — Wed, 08 Jul 2026 13:26:43 +0000

I have 15 years experience working in high tech electronics manufacturing (for a global firm with factories around the world, including in the US and China) and can tell you the most difficult thing to manage is the supply chain. For time as much as cost, but if you have to ship a $.10 component 6000 miles to get it to a factory, it's pretty dang inefficient.

That said, you're right about cultivating new experience in younger generations. It's probably not surprising to hear that the majority of factory workers in the US are GenX or older.

New comment by eitally in "Winamp Skin Museum"

eitally — Thu, 02 Jul 2026 15:36:13 +0000

This was also true on the comms apps sides. Back in the days of ICQ & Jabber, there was a huge ecosystem of chat apps, many of which (like my favorite, Miranda IM -- which has been revived for Windows as Miranda NG: https://miranda-ng.org/about/) were also fully customizable.

New comment by eitally in "Mag 7 starting to underperform [pdf]"

eitally — Mon, 29 Jun 2026 15:54:12 +0000

My perception is that the market lumps companies together into sectors whether they have similar business models or not. When one is punished, other associated firms tend to be, too. You see this in any industry (not as a guaranteed rule, but in general).

New comment by eitally in "Dostoyevsky isn't difficult"

eitally — Thu, 25 Jun 2026 05:32:24 +0000

I think a lot of the deeper enjoyment of literature comes when one has a sufficient understanding of the relevant culture & context and can adequately bring the characters to life. I studied Latin for 5 years (and Spanish for 3, Portuguese for 2 and German for 1), and I can tell you the immersion into Roman (and Roman Empire) culture absolutely made reading everything from Homer to Herodotus to Augustine to Seutonius to Cicero, Catullus or Ovid far more engaging than if I'd picked up any of these authors without context.

New comment by eitally in "How Madrid built its metro cheaply (2024)"

eitally — Thu, 18 Jun 2026 13:48:47 +0000

100% agree on all points!

And moreover, CHRSRA, while it is certainly delayed and over budget, appears to be a fairly well-run program with delays due primarily to external issues created by state & local government. It doesn't help that it receives nearly universal bad press, but I think that'll change by next year as they start laying track [connecting the stations and above grade crossings that have been long built].