... keep coming back to the same question with AI phone and operations agents: what should be disclosed before the agent is allowed to do anything that affects another person or system? I do work in this area, so I am biased toward over-communicating. My current instinct is that the minimum should be explicit identity, whether the conversation is being recorded or summarized, what actions the agent is allowed to take, when a human can step in, and some kind ... ... reversible low-risk tasks, but anything that creates a financial commitment, changes identity details, or sends a message the user would not want quoted later should force explicit confirmation. Then the receipt becomes the audit layer: what the agent disclosed

... standard editing. If you deploy emotion recognition or biometric categorisation, you have to tell the people exposed to it. Notably you don't have to explain the purpose just the operation. If you deploy a system generating deepfakes, you disclose. Same for AI-generated text published to inform the public on matters of public interest unless it went through real human review with someone holding editorial responsibility. Spell-check doesn't count. All of it has to be clear, distinguishable ...

... fuzzy on purpose, and fuzzy-on-purpose is exactly where indie products live. If you're in scope, here's what SB 243 actually requires, and I want to make the case that none of it is unreasonable. One: disclose that it's AI when a reasonable person might think they're talking to a human clear and conspicuous, not buried in a terms-of-service page. Two: have a real protocol for self-harm and suicidal ideation ...

... users , because the coverage of this was all about the parts that got delayed . Article 50 of the EU AI Act applies as of 2 August 2026 . In plain terms: An AI system that interacts with people has to disclose that it's AI (unless it's obvious from context) that's your support chatbot, your onboarding assistant. Synthetic media images, audio, video has to be labelled as artificially generated. AI-generated text published to inform the public needs disclosure ...

... actually failed, spanning 22+ countries and industries from fintech to biotech to retail. Each entry includes funding history, investors, a sourced explanation of what actually went wrong, a timeline, and a plain-language lesson, built to be a genuine alternative to expensive tools like Crunchbase that treat failure as a footnote rather than the actual subject. Every entry is manually researched and verified against public reporting, not scraped, and where a figure was never publicly disclosed

... shipped work where 90%+ of the code came from prompting an agent, with me mostly reviewing, testing, and directing the architecture. Never explicitly disclosed that to the client, they just know it got done fast, works well, and stayed in budget. Is that dishonest, or is it no different from using a framework, a library, or a junior dev and not itemizing every line's origin? Nobody used to ask what percentage of a website a developer typed themselves versus

... which is bounded, and cannot fire when nobody started anything. Your last seen sweeper is already the mechanism. It is sweeping into the wrong bucket rather than lacking the signal. Comment from Dale Mooney(@dalemooney): @aadilghani   The gate versus leash distinction is right, and sharper than what I wrote. Blast radius is knowable before the call. Time to notice isn't knowable at that moment at all, so using it as an approval input was me collapsing two different ... ... narrow the action (for example, create a draft instead of sending) or fail closed with enough context for the next human to resolve it quickly. Comment from Aria(@ariajinling): I agree: degradation has to be an explicit, author-authored alternative

... user and any friends they add. The free app, which offers some paid features, has about 4.3 million monthly active users, according to Sensor Tower data. The previously unreported acquisition closed this spring, Midjourney said. It declined to disclose financial terms. This is big: CEO @banu_guler is now Chief Design Officer at Midjourney. Comment from Varun S.(@varuns): Wonder what the angle is here. Looking forward to the shifts. Comment from Oluwole Zacheus(@emmadagold): This is a really interesting ...

... moved and tmux reattached elsewhere are the same failure underneath: the identity you resolved against stopped being stable, and nothing told you. Which is the argument I have been making all week about session memory generally, so I should disclose the obvious. I hunted Atuin on here today, which is the shell history side of exactly this: history that keeps the directory, the exit code and the duration instead of a flat file that forgets which session it belonged ...

... animation and 8 seconds of silence. Our engineer thought it was a bug when he first saw the spec. "You want me to add... a delay?" Yes. On purpose. Here's why. We ran an experiment: instant AI response vs. an 8-second pause before the same response. The content was identical. The only difference was time. Users who got the pause rated the AI's reflections 40% more helpful. Same words. Same insights. They just felt more considered ... ... honestly surprised us too -- we expected maybe 10-15%. It made us rethink a lot of our other interaction patterns. Comment from Mona Truong(@monatruong_murror): @alieksia   Really sharp question. We haven't run a controlled test on disclosed

... think "salesy" isn't really about word choice it's about whether you're answering something someone asked, or interrupting something they weren't asking about. Curious if others have found the line shifts once you have actual users vs. pre-launch does it get easier to talk about the product without it feeling like a pitch once there's real usage/feedback to reference instead of just claims? Comment from ivory xu(@ivory_xuxuxu): @suzychase  Extremely helpful, thank ... ... pushy or too salesy is simple: lead with the problem They see, position yourself as someone who understands it well, ask their expertise about solving the problem and then offer your product as a better solution or cost effective alternative

... play out there? Comment from Mira Charkawi(@mira_charkawi): I'm currently using phantombuster, gojiberry & Clay. Comment from Moeez(@moeez1): @sudipta_biswas4  Fair question. From what we ve seen, people don t really react to AI vs human , they react to whether the call is relevant and well-timed. And yes, if someone asks, it s always disclosed

... questions likes that under 10 seconds, end to end, presenting you answer, charts (if needed), all data used and a non-technical explanation what is what (if you drill down to it). When we do it, we don't disclose your data to LLM, while at the same time still using excellent coding agents in realtime to build, run and show you a one-off application to analyze and answer your question, all in under 10 seconds! Under the hood ...

... funding, I had to automate everything, therefore I relied on a cross environment team of AI assistants, that I, painfully and traumatically, intuitively guided to build deterministically. Since patent is pending, I cannot reveal too much, but I have disclosed enough for technical people to understand the architecture. You can also feed my announcement landing page URL to an AI assistant, so it could explain the technology ...

... actual insight, and it's bigger than it looks: making the notification say why it's waiting is what turns a ping into a triage. "Done" makes you tab back to find out what "done" means. "Needs approval vs finished vs

... economics just don't math. A few things I'm chewing on: Is usage-based or credit-based pricing now inevitable for any AI product with an agent layer? Anthropic claims Max 20x gives 20x the usage, but the disclosed numbers work out closer to 6x on Claude Code hours. Are we entering an era where "20x" means tokens, not outcomes? Cursor got dragged for communicating their pricing change poorly. Anthropic seems to be getting ahead of it with ...

... handle information. Comment from Farhad Asbaghipour(@farhad_asbaghipour): I d probably drop immediately trust is gone at that point. It s interesting how many modern workflows are becoming record first, ask later. Curious how others balance efficiency vs consent in client communication? Comment from Farhad Asbaghipour(@farhad_asbaghipour): @rohanrecommends  That s a really strong point once trust shifts, everything becomes transactional. I ve actually started seeing clients hesitate more, even before calls now. Feels like people are becoming more ... ... without acknowledging it beforehand is quite disrespectful in my opinion. In addition to trust, decency that comes with being forthright was lost. Definitely not the best way to establish a new relationship. Now, I wait to see if they disclose

... answer first. If they find it immediately, they cite it. The structure makes extraction easy. Step 2: The Comparison Page Template Comparison pages drove 40% of our citation growth. Here's our exact template: Title: [Brand X] vs [Brand Y]: Which is better for [use case]? Quick verdict (100 words) Who wins and why One sentence summary Feature comparison table Feature Brand X Brand Y Feature A Feature B Feature C Price XX YY Detailed breakdown (by feature) How each ... ... margin of error. Result: This block was cited verbatim by Perplexity within two weeks. Step 5: The 90-Day Update Cycle We analyzed citation decay across 500 pages. The drop-off is steep: Months Since Update Citations (vs

... think we all lie every day (even when we say "Good night" to someone, and we don't even have to mean it sincerely). We also lie on our resumes , we don't fully disclose everything on our tax returns , and we sometimes fake income . And that's what I want to talk about... Cluely CEO Roy Lee admitted that the $7M ARR figure he gave TechCrunch last year was fabricated , calling it the only thing he's "blatantly" lied ...

... loved one s voice using recordings that were fully consented to in advance. It is not about replacing someone or simulating the dead. It is about giving people a voice when silence becomes unbearable. Every call is disclosed. Every voice is permission-based. Every message is controlled by the person receiving it. Some people use it to say goodbye. Some to feel steady on hard days. Some just to remember what it felt like to be known ...

... with personal and work AI tools available everywhere (desktop, mobile, and now keyboards) For most people developers, writers, PMs, operators voice is already better than typing for a large portion of daily work. Snapshot: Pricing & Platform: Funding (publicly disclosed where available): Wispr: $81M+ total, including Series A and additional round in 2025 Aqua Voice: YC-backed; at least $500K+ publicly noted Willow Voice: YC-backed; thus also at least $500k raised Others : no public funding data; best understood ...

... infostealer-leaked credentials are uploaded every day. There are Telegram channels, dark-web shops, surface-web leak forums, private groups, and automated log dumping bots that publish millions of fresh credentials daily not months later, not after a company discloses a breach, but hours after a device gets infected. These logs usually include: Email accounts Social media logins Cloud access keys Developer tokens Session cookies Gaming accounts Crypto exchange credentials And more all stolen directly from infected devices Traditional breach ...

... bigger, but it pays off since automated regression is much quicker than manual testing . Comment from Mayas Haddad(@mayashaddad): That's an option worth considering. Although, for obvious ethical reasons, the non organic nature of the test should be disclosed by the influencer Comment from Shabnam Katoch(@shabb_katoch): Paying can get quick testers, but nothing beats organic curiosity. If someone wants to test your product, the feedback is usually way more honest. Comment from Mayas Haddad(@mayashaddad): @busmark ... ... value comes from the creator s ability to give authentic feedback while also exposing the product to the right audience. The compensation usually depends on factors like audience size, niche relevance, and the level of effort required (just testing vs

... anyone can just drop in a video and get instant insights all at $1 a minute...bulk runs get even friendlier on the pocket! Comment from Atique Bandukwala(@atique_bandukwala1): @busmark_w_nika  lol...i wish i could disclose but gotta respect them not wanting to do it and hence Comment from Atique Bandukwala(@atique_bandukwala1): @busmark_w_nika ...We ve trained our machine on 4.4 million videos spanning 50 geographies, 250+ languages, and more than 5,000 contextual ... ... metrics. You get detailed reports plus the ability to chat with video via our AI agent, Pixi. Pixi can: 1. Suggest how to rewrite a script to maximize viewership. 2. Predict how different audiences will respond. 3. Even help competitors

... wanted to commercialize his open source project, I believe he would be able to secure investment. Talent + interest + opportunity best wishes for these children. Comment from Andrew Jameson(@andrew_jameson111): @busmark_w_nika  I think I can't disclose his information casually, especially when he is still a minor. I can only say that he is the child of a former colleague and we met at a department dinner. Comment from Nika(@busmark_w_nika): @nitin_joshi ...