Top 26 stories from Hacker News. Top 10 include comment highlights. Compiled at 20:11 UTC.
1310 points by MaxLeiter · 569 comments
What HN said:
kimjune01: In the days when Sussman was a novice, Minsky once came to him as he sat hacking at the PDP-6. "What are you doing?", asked Minsky. "I am training a randomly wired neural net to play Tic-tac-toe", Sussman replied. "Why is the net wired randomly?", asked Minsky.
sumitkumar: The weights start with a random manifold. The training takes data and shapes the manifold, weight by weight, in many cycles. Once the training is the done manifold is fixed. When a new inference has to be done the query(q) is projected in the manifold space.
Planktonne: The original story is an original work made by a human consciousness exploring how it might be different from other forms of consciousness. This one is a pastiche made by a human consciousness borrowing extremely heavily from another human consciousness justifying why something e...
664 points by littlexsparkee · 621 comments
The percentage of failing grades in multiple UC Berkeley computer science classes in spring 2026 is significantly higher than past semesters and marks a departure from the department’s grading guidelines.
What HN said:
camelmel: I have some sympathy for these kids. If LLMs were around when I was a student, I would've also used them to "speed up" my homework assignments then proceed to fail all my tests. Now I work mostly with PhDs who were at the top of every academic environment they've ever been in.
somenameforme: The likely 'real' reason is hidden in one paragraph within the article and has nothing to do with the implication of the eye-catching title: "Both Garcia and Ranade have joined more than 1,300 UC faculty in signing a petition calling for the reinstatement of ACT and SAT standardi...
jventura: CS Professor here: just yesterday I did the discussion of a course projects' (Parallel Computing), and one of the three groups that I did yesterday have clearly gone the ChatGPT way. They couldn't even understand the choices the LLM made regarding the architecture, etc.
rahimnathwani: The grade data come from https://berkeleytime.com/grades I was worried they may have cherrypicked courses that support their chosen narrative. So I plotted the % of F grades (red line) for all CS courses still offered, and sorted the chart in descending order of the # grades give...
478 points by coloneltcb · 228 comments
VoidZero, the team behind Vite, Vitest, Rolldown, Oxc, and Vite+, is joining Cloudflare. Vite stays open source, vendor-agnostic, and built for everyone.
What HN said:
valgaze: "Vue.js: JavaScript MVVM made simple (vuejs.org)" February 3, 2014: https://news.ycombinator.com/item?id=7169288 Evidently Evan You was an Art History + Studio Art and major and at Parsons School he had to pick up javascript to quickly show his work.
yuppiepuppie: So is the business model of these projects - 1. build a popular dev tool 2. aquire funding 3. hire great talent 4. pray for an aqui-hire that justifies the initial funding amount I wonder how the initial investors feel about the aqui-hire path...
bluelightning2k: The reason this is worth it to CloudFlare is it will cause AI to recommend them more. The agents already reach for Vite. When they reach for Vite it's very logical they will default to CloudFlare after. (Much like they will guide users to setup Vercel for NextJS).
olingern: These acquisition announcements always leave me uneasy. There’s a lot of hand waving, “nothing will change and our roadmap will stay the same!” but we can all do basic math and understand that’s not how business works.
388 points by mooreds · 155 comments
What HN said:
jonathanlydall: I only realized in my 30s that I had been tying my shoelaces wrong my whole life and a super minor change in my method has changed them from coming undone multiple times per day (unless double knotted), to instead staying tied the whole day with just a standard shoelace knot [0]...
lee_ars: Stumbling across Ian's site almost two decades ago was kinda-sorta life-changing, because I'd been tying the "granny knot" my whole life and had to resort to double-knotting to keep my damn shoes tied. Ditched the granny knot for the Ian's Secure Knot (https://www.fieggen.
nunez: If you love this, you'll love learning about tying huarache running sandals: - http://borntorun.org/shop/howtotie.html - https://xeroshoes.eu/pages/tarahumara-sandals - https://importantbutnotatall.wordpress.com/2011/08/22/huarac... Infinitely many ways to tie these.
alper: I use Ian's (Fast) Knot and that's good enough for me. https://www.fieggen.com/shoelace/ianknot.htm
366 points by fidotron · 110 comments
What HN said:
everdrive: I always enjoyed the first half of Persepolis. Told from Satrapi's perspective, it was a very relatable story about a young child who was swept up by the world events around her, and tried to rebel in very normal, child-like ways.
tetrisgm: Poor woman. Somehow despite growing through hardships, it’s the loss of her husband that broke her. May she be at peace now, and her work cherished.
baby: I read all the persepolis comics a long time ago and to my memory it was the first time I cried reading a comic. A beautiful work of art. I would recommend to anyone reading this comment to order the first book.
letsdelta: I will always be grateful to her for so touchingly letting me in to her life, in a special part of the world, at a unique time of it’s existence
356 points by jc4p · 191 comments
As a part of my work I do security research for various apps and websites. I wanted to see if LLMs could reproduce a common class of exploits I
What HN said:
SOLAR_FIELDS: One interesting takeaway is the low score on Anthropic models from this benchmark. It’s not because of capability, it’s because Anthropic’s guardrails prevented it from solving the problem. I noticed with each model release Anthropic constrains the model more security wise.
mariopt: The methodoly used is quite naive. I've used glm 5.1 on fairly advanced crackme challenges (example: https://crackmes.one/crackme/698f40f1e2ba6023bfacaa82), and to my suprise it was able to patch binaries, doing runtime analysis, bypassing anti debug techniques, etc.
dwa3592: Nice exercise. Couple things: - I think the exercise was inconclusive for Claude and Gemini because they hardly tried to solve the task at hand. So the scores don't mean much.
mynameisvlad: It seems harsh to critique guardrails and take them into account in the scoring when GPT-5.5 seems to have been explicitly whitelisted to remove most of said guardrails. A more fair comparison would be a vanilla GPT account.
311 points by speckx · 272 comments
Wind and solar generated more electricity than gas globally for the first time ever in April 2026, according to new Ember data.
What HN said:
scope2093: We installed roof solar (10kW panels + 8kW hybrid inverter + 32kWh battery + planning/execution) last October for 11k euro. After all the math, our "investment" would pay off in approx 8-10 years (at current electricity prices).
whiteCaps: I just upgraded the solar system at my family's off-grid cabin. It's incredible how much battery technology has improved over the last 10 years. Everyone is getting tired of me checking the panel to see how many watts we're bringing in.
erelong: I know some people who are adamantly against solar and wind (personally I like both but I can see some shortcomings - for example I have heard that ai datacenters are using gas at times because of its flexibility) So what are some of the best talking points to "sell" solar and wi...
mtmickush: This is exciting news but the term power here should really be replaced with electricity which is clarified early on in the article. Electricity only accounts for roughly 20-25% of all power / energy used and the vast majority of the remaining 75% is fueled by gas (cars, ships, h...
224 points by rozumem · 210 comments
What HN said:
epistasis: NYTimes is predatory on subscriptions. Over my long lifetime I've subscribed twice, and regretted it both times with intensity. Any place that allows easy instantaneous subscription by a simple web form, but makes you call and talk to a person during limited business hours for ca...
devindotcom: It's not just them. Yeah, this is bad, but I get tons of unsolicited messages from any company I establish a basic relationship with. Every interaction I have with a store or site signs me up for some promotional thing, which I unsubscribe from immediately, only to find it's one...
ethagnawl: The UX anti-pattern of theirs which really grinds my gears is the "Continue reading in the app -- it's better." modal which appears when reading articles on the web. There does not seem to be a way to permanently opt out of it.
ilamont: If you think that's bad, 5 years ago you had to call someone on the phone to cancel NYT subscriptions (the boiler room retention script always gave you an option to extend at the cheaper rate, but it was a pain to have to go through the motions).
180 points by BrunoBernardino · 175 comments
Search privately and without ads using Uruky, the private search engine.
What HN said:
evilmonkey19: I really like the idea and that it's eu-made a love it. A fee things I see with kagi which are useful and improvements: - Hire a UI/UX person NOW! My parents and gf like using google and kagi because are easy to use.
maxloh: Could you consider Delayed Open Source Publication (DOSP) [1]? With that, you make the source code open-source after a specific period of time (such as 5 years). This way, your customers still get the source code and can use it freely, but you don't have to worry about competitio...
alex7o: I live in the EU and I use kagi, but I think you fail to understand why kagi is good and useful. In the end of the day I need a search engine to find stuff and kagi is better than google for the things I and my AI agents are searching for.
axegon_: I'd be more than willing to subscribe and support the project BUT, I need to address the elephant in the room: The reason why I'm against Kagi is the fact that they use Yandex(be it only for images according to their own words) and I'm sure as hell refusing to give a single cent...
178 points by tosh · 47 comments
What HN said:
Lammy: Big fan of this thing. It's one of my favorite places to take friends who visit the Bay Area. Something that's not mentioned in the article is that the building they occupy is a former warehouse of Marinship, a World War Ⅱ shipyard that made Liberty Ships and T2 oil tankers used...
LucasLanglois: If you don't know Tom Scott, he has done a great video 4mn vide on the model where you can see it in action: https://www.youtube.com/watch?v=i70wkxmumAw
hirpslop: The bay’s first LLM (large liquid model) was invented in the 1950s.
WillAdams: It's a shame that there isn't a series of articles on such models --- saw the Chesapeake Bay model (mentioned in a footnote) on a field trip when I was much younger (and it was still in active use for research I believe, yes, as my kids constantly tell me, I'm old).
AI/ML
Business/Tech
Other
Stories and comments sourced from Hacker News public API. Not affiliated with Y Combinator or Hacker News.