I'll be a proud daughter, my father built this and Baruch hashem shidduchim were already made! https://zuug.app/
8:23 PM
Excited to meet you all at the AI conference this Tuesday BeH!
5:16 PM
These are not even all the sessions! 🤯 Over 600 people expected BeH.
5:16 PMWhat is one wants to join many of the sessions of different tiers in the same time, will there will be post recording of those?
5:22 PMYup! Anyone who has a ticket will receive all 14 recorded sessions sometime after the event BeH.
5:22 PM
Due to popular demand for the AI Beginner Track, we’ve taken an even larger ballroom for the Beginner sessions at Tuesday’s PracticalAIconference.com in Staten Island. We’ve made some more Beginner tickets available as they were previously sold out. If you already bought an intermediate ticket after the beginner was sold out, please email [email protected] your first and last name, email address and that you’re requesting Beginner track. If you do not have a ticket yet and the website is now showing it’s sold out, I apologize as that means we officially sold out the larger ballroom as well. Don’t worry though: the conference sessions will be available for purchase in the weeks following this historic event BeH. 😊
9:42 PMDan Peguine built a custom screen-free voice messaging box for his kids to exchange WhatsApp audio messages directly with grandparents, featuring a large blue button for recording, ringtone alerts for incoming messages, message queuing, and scheduled quiet hours. The device was constructed following a detailed one-shot building manual generated by Fable AI, evolving from an initial idea posted weeks earlier into a fully functional two-way system that includes read receipts via WhatsApp reactions.
9:36 PMDan writes: I gave my kids a way to chat with their grandparents via WhatsApp. But without screens! This box lives in their bedroom, they press the massive blue button to record a voice message, and the message is sent to a group with both grandparents. It’s two-way so when the grandparents send a message back, there’s a ringtone (that we chose together) and the kids run to their room to listen. The grandparents even get a “listened” notification with a small headphones WhatsApp reaction. When the kids are away or asleep, the messages queue until their return. There are obviously quiet hours at night. My son and I built this over the past few weeks, iterating on the UX. The kids and grandparents absolutely love it! I love that they are in direct contact without me in the middle and without screens.
9:36 PM
Here’s a link to join a beta list or learn more: https://danpeg.ai/message-box
9:36 PM
Great use of AI video to promote no phone usage when talking to the One above. Video created by Mint Media. Watch: https://youtu.be/swLWZ572dAM Also, The Power of a Million is calling on Jews worldwide to commit to not talking during davening and be entered to win a $1,000,000 to benefit their shul: https://Ktyt.org/en/lcPOD
9:59 PM
And now is good a time as any to remind you about the terrifying truth about the device in your hand right now: https://youtu.be/vnGxdROljf8
10:00 PMI just want to say that ive been using gemini notebook for the first time , since seeing it at the conference. Its incredible. Im using it to help me prepare for multiple college courses , and so far it has produced for me flash cards , interactive quizzes , and mind maps. I probably havent even scratched the surface of what I can do, but wow. Thank you
12:58 AMI'm just letting you know I've been working on a project for the past month creating a school pamphlet for parents. Just a few of the tips that were shared about chat work versus chat regular, about starting new prompts, about how to prompt, got me to the finish line today. Finally after a month of working on this project and dreading it because it just wasn't coming to an end, I finally got to the finish line one day later. Thank you.
1:04 AM
Google Notebook is an incredible tool: https://notebooklm.google.com/
1:05 AMTell him to try taking a complicated subject and turning it into a podcast in notebook. You can even "call in" with questions.
1:06 AMWild stuff.
1:06 AMCall in how??
1:06 AM
Wow
1:06 AMIt's beyond cool
1:06 AMWe had some written testimony from grandparents and great grandparents who survived the Holocaust. We also found further history regarding a great great great great great grandfather who had made Aliyah. I was able to upload everything to NotebookLM and create a podcast geared towards kids so that my kids could get an overview without having to read more boring pages from the interviews. It was also able to make an easy to read family tree. It is also one of my favorite tools!
1:52 AMCute! I actually had fable write a full book based on newspaper articles, speeches my grandfather gave and an interview he gave with the Holocaust museum.
1:57 AMSame. Made a podcast of my grandparents escape from nazi Germany. Script from a German book that are translated through Claude and uploaded to notebook llm for a podcast
2:29 AMI took the study material for an exam I was taking into NotebookLM. Each chapter of the study material I put into a separate NotebookLM, and I had it make a podcast for each chapter so that I could study them by listening.
2:40 AM
New user feedback option on LinkedIn posts. 😂
12:06 AMAlex Finn: The greatest AI skill you can learn right now is Reverse Prompting It’s simple: ChatGPT5.6, Fable, and Grok 4.5 are now smarter than humans. That means you need to have your agent prompting YOU more than you prompt it You have your AI ask YOU questions to see what it can do for you Makes every AI agent you use so much more powerful An exercise you can do right now: Open up any AI/OpenClaw/Hermes. Brain dump everything about yourself and your career/goals/ambitions. Then use this reverse prompt: "Based on what you know about me and my goals, what is more information I can provide to you in order for you to be able to help me achieve my goals faster and take as much off my plate as possible" Then once you enter that, prompt this: "What tasks can you do for me right now to get us closer to our ambitions and goals?" Guarantee you come up with 100x more things to do with your AI than you thought of before The more questions you ask your AI, the more you'll learn and the more you'll get done. You don’t prompt. You give information, then allow your agent to guide you to somewhere interesting. This is how you use AI in 2026
1:20 AMI uploaded my booklist that I’ve read over the past few years to ChatGPT. (With a 1-10 ranking as well) I set up a daily automation in ChatGPT that brings lessons from books on topics I am working on that day. Great way to reinforce what you learned from the books
3:27 PMI programmed my phone to automatically go to a custom do not disturb mode when I arrive at shul
8:08 PMI even have a custom wallpaper with my Siddur to remind me to daven when I pick up my phone.
8:08 PMVery easy to set up and helps eliminate distractions
8:08 PMYes, leaving in car or locker is even better. :) Good start for those this can help BeH.
8:08 PMI did this as well. The problem I encountered was it goes by that address so driving by automatically set it off ( and I could be on a call ) bc it doesn’t differentiate between actually entering the building vs the vicinity of the address .
8:13 PMEver since Israel missed the Hamas leadership in Qatar because they left their phones behind to go pray I don’t bring my phone into shul
8:13 PM
lol. Thus concludes this week’s tips and tricks. ✌️
9:07 PM
Hey Eli, I bought the Kavana pouch which blocks wireless signals when you're in shul. It's a little embarrassing to have on your table and I know I could just power off my phone, but this works for me for some reason!
1:12 AMI'm actually running the iOS 27 developer beta on my phone. With the new update, you'll be able to easily create shortcuts with natural language. You can easily say, "Set up this Do Not Disturb mode at a certain time when I'm at a certain place, so it doesn't go off automatically."
1:12 AMThe first ever Practical AI conference was electric. ⚡ Over 500 attendees from more than 250 businesses came to NY to find out what’s real (and what’s not) in AI right now. Building custom agents. Getting employees to actually adopt it. Which tools are worth paying for. Where all of this is headed. At Maigrate, we build secure specialized AI agents and get teams to use them successfully. It’s exciting to see that the world of tomorrow is today.
4:00 AMA client of mine needed a list of around 20 dell computers with their status of their warranty and the basic information etc of the product so instead I should go on dells website and put in everything individually and then copy and paste everything on an Excel sheet I got Claude cowork to do it for me
3:01 AMBy the last 3 dell blocked it
3:01 AMHey everyone, Great meeting you all at AI Migrate. The follow-up guide I promised is attached. It is 7 steps, in order: get Claude set up right, bring your ChatGPT memory over, teach it who you are, connect your email, build your tone skill, record your first workflow as a skill, and find your five hours. Every prompt is in there ready to copy and paste. One warning before you start. If a step gives you a bad output, do not write it off. Stay in the same chat and let it ask you questions. That one habit is the difference between people who get value out of this and people who quit.
5:01 AMAI-Migrate Follow-Up Guide Hershey-Goldberger
5:01 AMThis guide is by Hershey Goldberger. He builds AI employees for businesses: office admin, customer intelligence, estimating, proposals, collections, outreach. hersheyg.com Stuck on a step? He opened a limited number of 15-minute slots for this group, on the house: https://calendly.com/hershey-g/15min
5:01 AM
Most frontier models have stale knowledge based off when they were trained and released. It used to be up to six months or a year, but see screenshot, opus 5.0 only has knowledge until May 2026. (Which in AI terms is like a year.) The user has to direct the models to use web research and find most up to date information.
3:46 PM
^ Important tip from Bryant Oberg of https://www.human-co-pilot.com/
3:46 PM
Grok is real time
6:02 PMI wanted to collaborate my kids school days off, so I asked ChatGPT to do that for me
10:32 PM
It’s always helpful when I see other people’s post so I figured I’d share as well
10:32 PMYou can upload this to Claude or really any models and ask it to put out in CSV so you could put it into a Google Calendar. I took my kids three calendars that were emailed to me in PDF uploaded them and had Claude just put it into one Google CSV and add a school May before the title of each event so I could see in my Google calendar all three school calendars consolidated
10:37 PM
I am not shocked that chatGPT can translate a full page of drone instructions from Hebrew to English. I am not shocked that it can graphically design it into a 1 page. I am shocked that we're living in a time when my 9yo oleh chadash on his own thinks to create this.
12:24 PMEnglish option on the top left.
1:56 PM
Hi Shani, how are you? I recently fulfilled a dream and built an app called “Cheshbon Nefesh.” It’s a personal tool that helps you pause for a few minutes each day, type or dictate what happened, and later look back, reflect on the week, and notice recurring themes. I’d really love for you to try it and share your honest feedback: https://cheshbon-nefesh-app.vercel.app/welcome?c=rabbanit-shani-taragin If you feel it could help others, please feel free to pass it along. Thank you! 🙏
1:56 PM
If you’re building AI agents, or just running into them in the wild, you’re probably asking: Which AI agents are quick and cheap to build and which take time and money to build? 🏰 Use this tool to get your answer: HowBigIsTheMoat.com (An AI agent is a program a person sets up with a goal and a set of tools. It works out which steps to take, uses the tools, and keeps going until the goal is done.) Product of Maigrate.com 🚀
3:36 PMUsing everything in my history (career decisions, health data, family structure, money system, stated fears, and the reversals I've made), project my current trajectory forward assuming no change. Identify the specific regrets I would most plausibly hold at 90, ranked by severity, where severity = irreversibility x compounding x how central it is to what I say I value. Do not use generic regret research. Use my own words and my own patterns as evidence. Then give me a sequenced action plan tied to real dates on my calendar, with pre-commitment devices that make follow-through independent of how motivated I feel that week. Tell me where I already know the answer and am avoiding it.
7:06 AMPrompt
7:06 AM
Google Search isn’t what it used to be. Story: https://www.theringer.com/2026/08/04/tech/google-search-ai-internet Businesses are investing more and more in AEO (Answer Engine Optimization), which is the practice of formatting content so AI tools and voice assistants like ChatGPT, Claude, and Google AI Overviews cite YOUR brand. If you’d like a contact successfully helping businesses do that, you can PM me.
4:24 PM
Here’s is a quick demo of TorahLM, a free tool designed to ground AI models in authentic Jewish texts and drastically reduce hallucinations when asking about Torah, Talmud, or Halacha. WATCH: https://youtu.be/bgGzA7ugtsQ When you ask standard AI models about Jewish topics, they can hallucinate or make up facts because they rely on generalized training memory. TorahLM solves this by giving your AI direct access to a massive, curated library of primary texts, Halachipedia articles, and shiurim; all formatted specifically for AI consumption. Instead of guessing, your AI can now read, search, and quote directly from the real sources, making verifiable citations the standard. TorahLM is free, ready to use, and seamlessly integrates into major assistants like ChatGPT, Claude, and Perplexity via a custom MCP connector, or through pre-loaded a Gemini Notebook.
1:49 PMAny questions, please contact the creator:
1:49 PM
Claude models will now have invisible watermarks embedded in ALL text, and ALL metadata attached to files. See: https://support.claude.com/en/articles/16266773-how-claude-marks-ai-generated-content
12:33 AM
Hey Guys, a good friend of mine built a Gematria site worth knowing about: 👉 www.TheGematria.com I've never seen anything like it before. You enter any number or Hebrew word, and it shows every word, phrase, or complete pasuk in Tanakh with that Gematria, all with sources and full tally views. It also has the full Tanakh with translation, full daily learning that updates each day, everything with audio, plus the Mishnah, both Talmuds and a Siddur. Installs on your phone like an app once you're logged in. Try searching your name or favorite number to see the full scope of its Gematria 🤯
2:49 PM
Of interest. If you have a Gmail account there are hackers making very similar accounts and emailing your Gmail account. Then when you try sending yourself an email it will often choose the email based on the first few letters and end up sending them the email instead of yourself. They can end up with private or any type of info that you shared yourself. Beware.
9:20 PMFor example an email like [email protected] they would make [email protected] or [email protected] would be [email protected]
9:20 PMWhat does ai recommend as actionable step
9:20 PMTo prevent
9:20 PMThis tactic is a variant of typosquatting and autocomplete poisoning, where attackers take advantage of how mail clients (and human memory) handle similar-looking email addresses. To prevent your email client from auto-suggesting these malicious addresses—and to secure your account against them—implement these targeted fixes: 1. Remove Bad Addresses from Google Contacts & Auto-Complete Google automatically saves addresses you've interacted with to your auto-complete directory. You need to purge the rogue email: * Go to Google Contacts while logged into your account. * Check under both Contacts and Other Contacts (in the left sidebar). "Other Contacts" is where Gmail silently stores every email address you've ever replied to or received mail from. * Search for variations of your email name (e.g., typos or added letters). * Click the three dots next to any suspicious contact and select Delete. 2. Turn Off "Create Contacts for Auto-Complete" Stop Gmail from automatically saving new, unfamiliar email addresses to your contacts list: * Open Gmail on a desktop browser. * Click the Settings gear in the top-right corner \rightarrow See all settings. * Under the General tab, scroll down to Create contacts for auto-complete. * Change the setting from "When I send a message to a new person..." to "I'll add contacts myself." * Scroll to the bottom and click Save Changes. 3. Add Yourself to Your Own Contacts Directory To ensure your actual address always takes priority in auto-complete predictions: * Go to Google Contacts and click Create contact. * Add your correct Gmail address and give it a clear contact name (e.g., "Me (My Real Email)" or your full name). * Whenever typing in the "To" field, always verify the avatar or full display name matches your verified contact entry before hitting send. 4. Create a Filter to Block Similar Domains/Typo Matches If you notice a specific spoofed domain or address pattern targeting you (such as @gma9l.com or [email protected]): * In Gmail, click the Show search options icon (the sliders in the search bar at the top). * In the From field, enter the malicious address or domain (e.g., @gma9l.com or [email protected]). * Click Create filter. * Check Delete it (or Never send it to Spam \rightarrow Apply the label if you want to isolate it for review), then click Create filter.
9:20 PMMoti Brody crushing the AI video space. You can contact him here:
10:57 PMAsk chatgpt "knowing me what are 20 useful apps i should check out that I might not know about"
10:58 PMI found this very helpful
10:58 PMHi, have a link to the ai chat I can share?
11:01 PMSending messages is only available in the app. Download
