Yury Shishkin

CEO & Founder of online retail through technology and AI.
🔴 IDRF has gone global! Meet us in Indonesia and Mexico. Get free tickets below

[Yury Shishkin @24ttl]

Hi! I'm Yura, the Founder&CEO of 24TTL, an online merchandising solution that shortens your customer's decision-making process. We help brands like Samsung, Sony, Pfizer, Bosch and P&G improve their shopping experiences. Follow for news and insights from the #ecommerce industry!

[Yury Shishkin @24ttl]

OpenAI just launched GPT-Live, a new voice mode for ChatGPT that feels like a real conversation.

You can interrupt it, and if you go quiet to think, it waits instead of talking over you. Hard questions get handed to GPT-5.5 in the background while it keeps chatting.

[OpenAI @OpenAI]

Introducing GPT-Live, a new generation of voice models for natural human-AI interaction.

Rolling out in ChatGPT starting today.

You’ll want to turn the sound on for this one.

[Yury Shishkin @24ttl]

2/ Will I trust an agent to pick the right product, arrive on time, handle returns, avoid fakes?

In a world of AI agents, you compete on trust in the platform, not just price or selection. Most underrated topic in ecom right now.

[Yury Shishkin @24ttl]

1/ In the Valley today I met Daniel Danker, EVP of AI at Walmart (2.1M employees, ~480M store visits a week).

His sharpest point on agentic commerce: the differentiator won't be AI. It'll be trust.

[Yury Shishkin @24ttl]

Meta just introduced Muse Image and Muse Video, the first media-generation models from Meta Superintelligence Labs. Muse Image is their most advanced image generator to date.

[AI at Meta @AIatMeta]

Introducing Muse Image and Muse Video, the first media generation models developed by Meta Superintelligence Labs.

Muse Image is our most advanced image generation model yet. It follows instructions faithfully, edits with precision, composes from multiple references, and draws on Instagram for social context. It also brings agentic tool use capabilities to image generation and integrates with Muse Spark.

You can try Muse Image in the Meta AI app and web, as well as in Instagram Stories and WhatsApp – starting in limited countries with more locations on the way.

Today we’re also previewing Muse Video, which is built upon the same pretraining base as Muse Image to deliver exceptional visual fidelity with native audio support.

Learn more about both models: https://t.co/QtKDPDZP5v

[Yury Shishkin @24ttl]

Claude Cowork is coming to mobile and web.
Hand Claude a task at your desk, then pick up the finished work from your phone. Close the laptop and it keeps going. Beta rolls out over the next few weeks, starting with the Max plan, more plans to follow.

[Claude @claudeai]

Claude Cowork is coming to mobile and web.

Hand Claude a task at your desk and pick up the finished work from your phone. Close the laptop and Claude keeps going.

Beta is rolling out over the next several weeks starting with the Max plan, with more plans to follow.

[Yury Shishkin @24ttl]

An open-source alternative to Claude Science just dropped: OpenScience.

Installs via npm, runs in your browser, takes a research goal to a finished report: reads papers, writes code, runs experiments, queries ~30 science databases.

Any model via your own keys. Apache 2.0. #AI

[Yury Shishkin @24ttl]

Romain Torres built a Claude skill for motion-style animation. Paste a reference image, Nano Banana 2 makes the frames, Seedance 2.0 animates it. All inside Claude via Arcads MCP. 100% automated.

[Romain Torres @rom1trs]

I built a Claude skill that makes motion-style animation.

> Paste a reference image
> It generates frames with Nano Banana 2
> and animates the video with Seedance 2.0

Everything runs in Claude through the Arcads MCP

Comment "Motion", and I'll send you the skill

[Yury Shishkin @24ttl]

Fable 5 is back. Pulled for 3 weeks over export controls, live again globally as of July 1 with a new safeguard that blocks the jailbreak in 99%+ of tries.

[Claude @claudeai]

Fable 5 is back.

[Yury Shishkin @24ttl]

ok Sonnet 5 is kinda wild.
it plans, drives browsers and terminals, runs solo like a proper agent.
this thing does what only the big expensive models could pull off a few months ago. ngl the mid-tier just ate.
Sonnet 5 W or nah?

[Claude @claudeai]

Introducing Claude Sonnet 5, our most agentic Sonnet yet.

It makes plans, uses tools like browsers and terminals, and runs autonomously at a level that just a few months ago required larger and more expensive models.

[Yury Shishkin @24ttl]

NotebookLM can now build a 60-second vertical video from your own files.

Upload an article, a book or your notes, and it returns a Reels-style clip explaining the concept. Text and audio were the old outputs, video is the third.

Ultra and Pro only, English sources, free later.

[Yury Shishkin @24ttl]

The demo is wild, but hundreds of warmed accounts is exactly what TikTok and IG keep purging. Curious how this survives a single trust-and-safety sweep

[Gaurav @gauravsbuilding]

Today we’re introducing Claude for Social Media Accounts.

Just enter your website and Fastlane creates hundreds of accounts on TikTok and Instagram and warms them up in your niche. All done by real humans.

This is seriously insane.

[Yury Shishkin @24ttl]

Soon you won't ask a friend for help, you'll ping their agent. Strange and inevitable

[0xDesigner @0xDesigner]

talk to a friend's codex

[Yury Shishkin @24ttl]

Microsoft just turned Copilot in Excel into a finance analyst.

Three upgrades: reusable Skills (DCF models, quarter close, board reports), six new data connectors (PitchBook, FactSet, S&P Global and more), and a Plan mode that shows every cell it will touch before editing. #AI

[Yury Shishkin @24ttl]

OpenAI just built its own chip.

With Broadcom it unveiled Jalapeño, a custom accelerator made only for LLM inference, designed from scratch in 9 months. OpenAI even used its own models to speed up the work.

First gen targets better performance per watt than top GPUs. #AI

[Yury Shishkin @24ttl]

AI virtual try-on is quietly becoming a retention engine for fashion.

In DressX's new report, shoppers who used try-on added to cart 3x more often and showed 7x higher retention. They don't just shop, they play, and the play pulls them back. Rewrites CAC and LTV. #ecommerce #AI

[Yury Shishkin @24ttl]

OpenAI's Codex quietly went open.

With the --oss flag you point it at local models like Ollama or LM Studio and run the whole coding agent on your own machine, no cloud, no subscription, no rate limits.

Set the default in config.toml via oss_provider. #AI #Codex

[Yury Shishkin @24ttl]

Walmart vs Amazon, two bets on AI shopping. Walmart brings its own agent Sparky into ChatGPT and Gemini and joins Google's open commerce protocol. Amazon keeps buying inside its own app with Buy for Me. Same goal: own the agent.

https://t.co/HNWm5wrZhK

[Yury Shishkin @24ttl]

Shopify just put store marketing on autopilot. Campaign Autopilot runs and optimizes campaigns across Facebook, Instagram, Shop and email inside the admin, within guardrails you set. Plus an AI sales associate in Inbox that answers buyers from your real catalog. #ecommerce

[Yury Shishkin @24ttl]

https://t.co/am8N0lBU63 just dropped GLM-5.2. A 1M token context window, 5x bigger than its last model, holds a whole codebase in one pass. Tops open-weight models on the intelligence index, level with closed models from Anthropic and OpenAI. MIT weights land next week. e

[Yury Shishkin @24ttl]

Amazon's Rufus AI is used by 250M shoppers, and it quietly changed selling. "Researched by AI" blocks now sit above your listing and cite outside blogs and videos first. One mention in a trade blog can beat your perfect listing. Keyword stuffing is dead. #amazon #ecommerce