theverge.com
TikTok will pay $400 million to settle DOJ child privacy lawsuithttps://www.theverge.com/tech/983531/tiktok-settle-doj-lawsuit-coppaThe US Department of Justice announced on Friday that TikTok will pay $400 million to settle a lawsuit filed in 2024 over allegedly violating the Children's Online Privacy Protection Act (COPPA). In the lawsuit, the DOJ alleged that TikTok collected data from children without notifying parents or obtaining consent and did not delete the accounts […]Aug 21, 2026 10:13 PM
cbsnews.com
TikTok agrees to pay $400 million to settle child privacy lawsuithttps://www.cbsnews.com/news/tiktok-settle-child-privacy-lawsuit/The Justice Department sued TikTok in 2024, alleging that it violated a U.S. law prohibiting the collection of kids' data.Aug 21, 2026 9:08 PM
github.com
Show HN: Zroar – Serialized Roaring Bitmaps in Zighttps://github.com/manishrjain/zroarzroar is a ground-up implementation of Roaring Bitmaps data structure in Zig. zroar stores both the keys and (array, bitmap) containers in a single flat byte buffer, making the in-memory representation equal to the on-disk or over-the-network representation, eliminating the serialization/deserialization step entirely.The design was originally aimed at systems which keep their posting lists on disk, but zroar performs faster than CRoaring even for purely in-memory ops, due to CPU cache locality.Against CRoaring 5.0's benchmarking suite (ported to Zig), zroar is faster in 339 out of 360 tests, being 2x-9x faster (geometric mean), and up to 600x faster on serialize/deserialize.zroar avoids complex mechanism (like adaptive radix trees), uses Zig native SIMD ops and is simpler. The main logic is written in ~2000 lines of code, while CRoaring's 64-bit bitmap codebase is over 17000 LOC.Not yet: By choice, zroar doesn't support run containers, and is 64-bit only.There are more details in the GAug 21, 2026 8:18 PM
techcrunch.com
Why is the DOJ investigating Andreessen Horowitz’s board seats?https://techcrunch.com/video/why-is-the-doj-investigating-andreessen-horowitzs-board-seats/Andreessen Horowitz has two partners sitting on the boards of companies that now compete with each other: Ben Horowitz at Databricks and Martin Casado at Fivetran. Nothing too scandalous on the surface, except the Department of Justice has reportedly been investigating the arrangement for almost a year, dusting off a 112-year-old antitrust law that’s rarely used against VCs. Board conflicts aren’t exactly new, and these companies weren’t necessarily direct competitors when a16z first invested […]Aug 21, 2026 4:53 PM
github.com
Show HN: Bounded GenAI metrics from Otel traces with O raw promptshttps://github.com/llm-measurement/otelcol-genai-sketchesHi friends,Author here. Like to present a project I've been working on - Otelcol-GenAI-Sketches, a Otel connector that works on genai/agent traces, and summarizes them (via sketches), to produce:() exact request counts/totals for tokens; () distinct users/prompts (estimates) (*) 'heavy' items in terms of tokensWhy does this exist: I'm a data scientist by profession, and needed to look at Otel traces, but soon ran into cardinality and portability issues, along with privacy issues (raw prompts being transported elsewhere, like for analysis in a notebook env is a no-no). I wrote a sketchlib first (llm-sketchkit) and then the connector for Otel collectors.To try it out, just:(i) git clone https://github.com/llm-measurement/otelcol-genai-sketches.gi... (ii) cd otelcol-genai-sketches (iii) export GENAI_SKETCH_SECRET="$(openssl rand -hex 32)" (iv) make example-up (v) profit!This would start a demo app that includes a Otel collector, prometheus server/db and a grafana dashboard.The repos contaAug 21, 2026 3:37 PM
tosynch.com
Show HN: Tosynch – reconnect with your friends and meet oftenhttps://tosynch.comI built this app to solve problem I had, and perhaps others too. I have been finding tough stay in touch with friends. Other than occasional hi/hellos, there is lack of meaningful interactions ongoing basis. Tosynch helps with connecting with existing friends, without calendar setup. It does one thing only and focuses on that. There is no hidden agenda of selling personal data for ads, nor any scheme around increasing likes. There is no discovery mechanism, and no public profiles. If we want to reduce our usage of phones and devices, we don't need another app to helps us constrained. We want to create more opportunities to meet people we already know. Tosynch helps increase those opportunities. It nudges after a chosen period, to reconnect and meet. Feedback appreciated.Aug 21, 2026 3:10 PM
bing.com
PlayStation 6 handheld not expected to stay under $600, warns leakerhttp://www.bing.com/news/apiclick.aspx?ref=FexRss&aid=&tid=6a88d65da8b6480ca2484a49f3b265ad&url=https%3A%2F%2Fwww.tweaktown.com%2Fnews%2F113194%2Fplaystation-6-handheld-not-expected-to-stay-under-dollars600-warns-leaker%2Findex.html&c=4414574365890945269&mkt=en-usProminent AMD leaker says a $600 PS6 handheld isn't possible right now, blaming LPDDR memory prices that have been squeezed by AI data center demand.Aug 21, 2026 12:50 PM

france24.com
Cybercriminals perfect 'social engineering': 'Each hack feeds the chances of the next one succeedinghttps://www.france24.com/en/video/20260821-cybercriminals-perfect-social-engineering-each-hack-feeds-the-chances-of-the-next-one-succeedingPeter O'Brien is pleased to welcome Simon McGarr, Lawyer with McGarr Solicitors in Dublin, and Managing Director of Data Compliance Europe. Cybersecurity is no longer simply a question of whether governments possess adequate technical defences, explains McGarr. AI is altering those risks at extraordinary speed, placing capabilities once associated with states or sophisticated criminal organizations within reach of much smaller actors. If threats evolve month by month, governments must focus on prevention: identifying vulnerabilities, adapting defenses and developing long-term institutional capacity before emerging threats become crises.Aug 21, 2026 10:04 AM

bbc.co.uk
Millennials have found it hard to buy homes – but things may be turning a cornerhttps://www.bbc.co.uk/news/articles/cgewlld498xo?at_medium=RSS&at_campaign=rssTwenty-somethings are much less likely to own a home than previous generations. But data suggests things may be improving.Aug 21, 2026 10:02 AM
techcrunch.com
Learn what VCs actually want, from a founder who’s raised $1Bhttps://techcrunch.com/video/learn-what-vcs-actually-want-from-a-founder-whos-raised-1b/Investors want founders who understand the financial reality of their business. Messy data, misunderstood metrics, or waiting until you’re nearly out of cash to start fundraising can cost founders leverage, valuation, and even a term sheet. In this episode of Build Mode, host Isabelle Johannessen sits down with Sasha Orloff, founder and CEO of Puzzle […]Aug 20, 2026 11:32 PM
basecompute.co
Show HN: Zero () friction local AI for Machttps://www.basecompute.co/localSuper excited to launch our new app Local today. What we’ve learned at Base Compute over the last months is that running AI directly on your laptop or workstation gives you maximum privacy and it’s free, but it’s also a massive headache to configure. So we’ve decided what matters is making the experience completely frictionless for users.Local analyses the hardware of your laptop, optimises the AI for it, and recommends the best models for your specific device.It let’s you do what you’re doing with cloud AI already, just for free and on your own machine: Chatting with PDF’s, Recording and summarising meetings, running coding agents...If you’re using Local in your office with colleagues, you can run it in “Office Mode”. The strongest computer in your office runs the AI and everyone can connect to it with their laptop. The data never leaves the office.It’s available for download on our website today, please try it out and let us know what you think!Aug 20, 2026 10:13 PM
news.ycombinator.com
Show HN: We Added Notion Calendar to *Superhuman for iMessagehttps://news.ycombinator.com/item?id=49378903Another update on Attention: after launching here last week, over 100 people used Attention!We followed up on feedback and we added a Notion Calendar-style calendar!---We realized pretty quickly when working in Attention (Superhuman for iMessage) that a lot of our conversations were about scheduling things.When we talk to Righthand users, iMessage is a lot of "can we find time to debug the issue live" or "I'll get back to you by this evening with the answer". Personally it's a lot of booking travel and dinners.Since we've been spending so much time in Attention, we decided to add a calendar to fix the issue. It looks like Notion Calendar, but it also supports a few new features like "fuzzy events".The onboarding is more involved and so we'll do 1:1s for a bit (comment below if you want to be in the cohort), but we would love any and all feedback!As before:- Privacy / Security: Your data is yours, fully local in LanceDB + SQLite, never leaving your machine unless you turn on the AI agenAug 20, 2026 7:17 PM
news.ycombinator.com
Are you good at AI, or just using it?https://news.ycombinator.com/item?id=49378057We’re working on a ladder for individual AI proficiency and would love feedback on both the levels and the definitions.L0 New: brand new to AI, or has not yet used it.L1 Chat: simple prompt-and-response use. Work is serial: ask, wait for an answer, then ask again.L2 Contextual Work: gives AI relevant documents, data, or workspace context so it can work within the actual artifact and produce a more useful result.L3 Orchestrate: coordinates multiple agents or AI roles across independent workstreams, with work that may review, challenge, compare, or build on other work. This is not just for engineering.L4 Automate: creates workflows that are triggered by business events and run without someone sitting at a laptop directing each step.L5 Loop: feeds the output of those workflows back into shared knowledge or a company brain, so future workflows improve over time.A few things I’d love your perspective on:Are these the right levels? Are any of the names unclear or overlapping? What observableAug 20, 2026 6:04 PM
github.com
Show HN: Pond – lossless archive for agent sessions in your own S3https://github.com/tenequm/pondI wanted all of my sessions preserved in one S3 storage dump I could access from anywhere. I couldn't find anything that could store sessions from multiple machines in one remote S3 bucket without running a database service anywhere, make them available as an MCP for my agent, and efficiently search through them. pond was born out of this, and out of my necessity to stop being locked to my laptop. I'm convinced that the sessions we generate with agents are one of the most precious assets we have and it still feels so odd to me that we throw them away like we couldn't care less.With pond you just point it at an S3 bucket or a local directory. Under the hood is in-process Lance. It also supports safe concurrent writes, so I just collect all my sessions from all my local machines, remote VMs and telegram agent setups in one bucket dir. Right now that is 14,861 sessions, 2.83M messages, 10.6 GiB, from 8 tools (Claude Code, Codex, opencode, pi, Claude Desktop, oh-my-pi and a few more).The oAug 20, 2026 4:04 PM
github.com
Launch HN: Vendo (YC S26) – Let users build features on top of your producthttps://github.com/runvendo/vendoHi HN, we’re Yousef & Nour, founders of Vendo (https://vendo.run). Vendo lets users create new features inside the software they already use. A user describes the dashboard, workflow, or small app they need, and Vendo builds it on top of the product’s existing data, API, and interface.Demo: https://www.youtube.com/watch?v=VdpHehY64lsWe built Vendo because every SaaS eventually faces the same problem: every customer needs something slightly different. One wants a new report and another needs a workflow that only makes sense for their team. These requests either sit on the roadmap, become one-off engineering work, or force the customer into spreadsheets and external tools. We wanted the user to be able to create the missing feature themselves, without leaving the product.Here is how it works:- npx vendo init reads the product's API surface, theme, routes, and more. These are used so that the apps Vendo creates (1) look on-brand and native and (2) have the ability to read data and performAug 20, 2026 3:29 PM

arstechnica.com
Reverse-lookup service exposed millions of photos of people’s faceshttps://arstechnica.com/gadgets/2026/08/reverse-lookup-service-exposed-millions-of-photos-of-peoples-faces/People-search tool ClarityCheck left database containing more than 9M image files exposed.Aug 20, 2026 1:29 PM
techcrunch.com
OpenAI seeks to one-up Anthropic with new customer privacy protectionshttps://techcrunch.com/2026/08/19/openai-seeks-to-one-up-anthropic-with-new-customer-privacy-protections/A competition is developing between OpenAI and Anthropic over who can provide the best privacy protections for enterprise customer data.Aug 19, 2026 10:10 PM

theverge.com
Nielsen is leaning more on wearables to hear what people are watchinghttps://www.theverge.com/entertainment/982358/nielsen-ratings-ppm-smart-watch-cowatching-big-dataIn order to beef up its ability to accurately measure viewership data in the streaming era, Nielsen is moving forward with a plan to use more information gathered from its partners' wearable devices.
Today, Nielsen announced "a number of key enhancements" that are being added to its data capturing process ahead of the upcoming fall TV season. As of August 31st, Nielsen's co-viewing data will now i...Aug 19, 2026 7:53 PM

techcrunch.com
Meet the startup helping Wall Street put a price on AI computehttps://techcrunch.com/video/meet-the-startup-helping-wall-street-put-a-price-on-ai-compute/The AI buildout shows no signs of slowing. And with hundreds of billions of dollars a year going into data centers and GPUs, compute has become the single biggest cost for anyone building AI products. But for all that spending, there still isn’t a straightforward way to put a price on compute — or for firms to hedge their exposure when the price changes. Silicon Data […]Aug 19, 2026 5:26 PM

fortune.com
‘Often flying blind’: Up to $246 million of taxpayer-funded CDC programs are largely unstaffedhttps://fortune.com/2026/08/19/often-flying-blind-up-to-246-million-of-taxpayer-funded-cdc-programs-are-largely-unstaffed/The Office on Smoking and Health, the epilepsy program, sickle cell disease data collection operation, the rape prevention unit: all zombies.Aug 19, 2026 5:00 PM