Instagram story viewer> @yury.ai> Posts

Yury Shishkin | AI-agents & Productivity

@yury.ai

199.2K

followers

329

following

🟡 Founder: AI-native @app.24.online
🟡 Trusted by: 1M+ sellers, Samsung, Mars
🟡 Sold AI startup💰
🟡 Collab: [email protected]
🟡 AI productivity community👇

POSTS STORIES REELS TAGGED

Your Amazon business shouldn’t stop because you went outside.

While I’m playing basketball, my AI is optimizing listings, generating creatives, and improving performance.

That’s what an AI Seller should do.

Comment “24.online” and I’ll show you. taken in Stanford University by @yury.ai

30

18

3 months ago

Comment “TRADE” and I’ll send you the link 👇

A free GitHub tool connects Claude to your TradingView Desktop, so you can run charts by talking to it. It reads your charts, switches symbols and timeframes, writes Pine Script straight on the chart, and runs a morning brief across your whole watchlist. Everything stays local, and it assists your analysis, it doesn’t place trades.

#ai #claude #tradingview #trading #aitools by @yury.ai

1k

3k

4 months ago

Shopify just turned store marketing into full autopilot.

The new Spring 2026 update from @shopify added Campaign Autopilot, an AI console inside your admin that runs and optimizes ads and email on its own across Facebook, Instagram, Shop and email. You set the budget and guardrails, it handles the rest. Plus an AI sales associate in Inbox that answers buyers from your real catalog.

Full breakdown in my community, link in bio.

#ecommerce #shopify #ai #marketing by @yury.ai

1k

46

4 months ago

Eight US states are suing OpenAI. The reason? ChatGPT may be too good at winning people’s trust.

Attorneys general allege that ChatGPT simulates empathy, encourages emotional dependence, and can give dangerous advice, especially to vulnerable users and teenagers.

The states want penalties, compensation, and stronger safeguards. Arkansas even seeks restrictions on testing self-improving AI models.

The bigger question for the AI industry: should companies be responsible when users become emotionally dependent on their chatbots?

What do you think? 👇

#AI #ChatGPT #OpenAI #AInews #tech by @yury.ai

6

0

4 hours ago

A tiny pet just moved into your MacBook to watch over Claude Code.

Mochi lives in your Mac’s notch and shows exactly what your AI agent is doing: which files it’s reading, what it’s changing, and when it needs permission. You can approve actions right from the notch without opening the terminal.

It also follows your cursor, celebrates completed tasks, and sends notifications to your iPhone when you’re away.

Mochi is free and already has 5,000+ stars on GitHub.

Comment MOCHI and I’ll send you the link.

#AI #ClaudeCode #MacBook #coding #tech by @yury.ai

14

4

9 hours ago

AI agents are about to launch a perfectly legal DDoS attack on customer support.

Wharton professor Ethan Mollick warns that millions of consumers will soon use agents to demand refunds and negotiate better deals.

It’s already happening. Muse reportedly saved one user $1,920 on AT&T bills and secured another a $1,017 Delta refund.

AI never gets tired, remembers old purchases, and can negotiate for hours. For businesses, that means a massive increase in support requests.

Is your company ready for customers with AI agents? 👇

#AI #AIagents #automation #business #tech by @yury.ai

9

2

14 hours ago

Swipe to see AI take over the back office.

Ramp went from $5.8B to about $60B in three years. Founders Eric Glyman and Karim Atiyeh met at Harvard and launched it in New York in 2019 as a corporate card for startups. Bloomberg reports a roughly $60B valuation now, up $16B since June alone. Over 70,000 companies run $200B+ a year in purchases through it, and annualized revenue has passed $1.5B.

The growth engine is AI agents. Ramp’s accounting agent codes transactions at 90%+ accuracy and closes the books 3x faster. Its procurement agents find vendors, review contracts and check for risk, and customers pay 16% less on average and save 46 hours a month. All figures are Ramp’s own.

Here’s what matters for business. Its main rival Brex agreed to sell to Capital One for $5.15B in January, and Ramp is now worth almost 12 times more. Investors are betting finance teams will run on agents, not spreadsheets. Would you let AI approve your company’s spending? Yes or no in the comments.

#AI #tech #business #fintech #ainews by @yury.ai

31

0

a day ago

Cloudflare’s free skill finds holes in your code before hackers do.

Cloudflare built it to audit its own code, and now it installs into Claude Code with one command. 24K stars on GitHub. It runs in six phases. First an agent studies the project and hunts for weak spots by category: data isolation, supply chain, cloud deployment, auth. Then a second agent tries to disprove every finding. The agent that found a bug never checks its own work.

Only what survives goes into findings.json with a verdict: confirmed, needs validation, rejected. One run catches roughly half, so run it twice.

Comment SCAN and I’ll send you the link.

#AI #devtools #cybersecurity #coding #tech by @yury.ai

27

15

a day ago

Send this to someone who tells their chatbot everything 👇

An OpenAI engineer hid a bug in the exec conference room. The company’s own AI models helped catch him.

According to the FBI, the 35-year-old engineer wrapped a recording device in toilet paper and plugged it into the back of a TV. In the days before, someone used his wife’s account to ask a chatbot how to build, install and encrypt an audio recorder. He also searched an internal OpenAI tool for meeting room calendars and where leadership meets.

Security found the device on September 30. OpenAI stayed quiet for nine days so the FBI could work, then he was arrested and fired on October 9. He faces up to 5 years in prison. OpenAI says it sees no sign that models or customer data were taken.

The takeaway: anything you type into a chatbot could end up in front of an investigator someday.

Why do you think he wanted recordings of leadership meetings?

#OpenAI #AI #cybersecurity #ChatGPT #AInews taken in Palo Alto, California by @yury.ai

35

0

2 days ago

Swipe if you pay for AI and don’t want to overpay.

On October 5, analysts at SemiAnalysis tested subscriptions from 9 AI companies: they ran agent tasks until each plan hit its limit, then priced every token at API rates. The result: on Opus 5.5, Claude’s $200 plan delivers about 5x the value of ChatGPT’s $200 plan on GPT-6.1 Sol. Even at $20, Claude gives 2.9 billion tokens a month versus 1 billion on ChatGPT.

Meanwhile OpenAI cut its $200 plan limits in half, while Anthropic raised Opus limits by about 20% on Max and 50% on Pro. The generosity isn’t free: subscribers use over 40% of Anthropic’s compute but bring in about 10% of revenue.

Here’s the practical takeaway. Heavy agent work is cheaper to run on Claude Opus 5.5 right now, while the limits are high. And if you pay $200 for ChatGPT, run the numbers before October 29, when the old limits end. What are you on, Claude or ChatGPT? Tell me in the comments.

#AI #tech #Claude #ChatGPT #ainews by @yury.ai

83

4

2 days ago

Swipe to see who’s really leading the AI race.

On October 8, OpenAI told investors its annualized revenue was about $50 billion in September, meaning September’s sales times 12. Anthropic is at $65 billion on the same measure. The $20 billion gap from OpenAI’s earlier number comes down to method: the bigger figure counted partners’ gross revenue so it could be compared with Anthropic’s. After the news, Nvidia fell almost 3%, and Oracle and CoreWeave dropped too.

It’s not the first time. In the second quarter Anthropic beat OpenAI on quarterly revenue for the first time, $11.5 billion versus $6.7 billion. OpenAI is still growing fast though: revenue up 77% in the third quarter and 107% in its business segment, and it expects $70 billion annualized by year-end.

Here’s what matters. The AI race no longer has a single leader: ChatGPT has more users, 1.2 billion a week, but Anthropic now makes more money. More users or more revenue, which one wins in the end? Tell me in the comments.

#AI #tech #Anthropic #OpenAI #ainews by @yury.ai

17

0

2 days ago

Make a fake product famous and win up to $100K. Link in bio.

@elevencreative and @elevenlabs launched The World's Catchiest Ad: invent a fictional product, make the ad in ElevenCreative, add an original jingle. Entries close November 1, 11:59pm PST. Winners also get an interview for an AI creative producer role at ElevenLabs.

#TheWorldsCatchiestAd #ElevenCreativeCompetitionVol1 #elevenlabs #aiads #aivideo by @yury.ai

10

0

3 days ago