LATEST AI NEWS
AI NEWS
Slack Launches Code Channels for AI Agents
AI NEWS
Linkdaze Puts an AI Meal Planner on the Kitchen Wall
HEALTH
FDA Says Generative AI Rules for Medical Devices Are Coming
SOCIAL MEDIA
LinkedIn's AI Slop Button Passes One Million Clicks
Google Discover Gets a Chatbot-Tuned Feed
Google Discover is adding a way to shape your feed by describing what you want in a chatbot style prompt. The tool tweaks your recommendations and remembers your preferences, rolling out in the Google app over the coming days.
Source: The Verge
AI NEWS
Slack Launches Code Channels for AI Agents
Slack has launched Slack Code, a new kind of channel where engineers and AI coding agents plan, write, and review software in the open rather than inside a private browser tab. Mention a coding agent on a real project and a dedicated code channel spins up, pulls in the right teammates, and shapes itself around the task at hand.
Inside a code channel, the agent can surface a code diff, a planning document, or a live HTML preview directly in the conversation, so the whole team watches the work take shape. People can catch a bug, flag a constraint, or redirect the agent mid task instead of discovering the result after it ships. When the job is done, the channel archives itself but stays searchable, acting like an audit log for later.
Slack Code inherits Slack's existing permissions and admin controls, so IT does not have to configure anything new, and high stakes changes can route to a person for a quick approval inside the channel. Slack says more than 70 percent of its own code channels open and close within a single day, moving from idea to merged pull request.
The launch arrives with a roster of partners. Teams can already work this way with Claude from Anthropic, Devin from Cognition, GitHub Copilot, and Vercel, with ChatGPT from OpenAI coming soon. Slack frames engineering as the starting point and says the same collaborative model will later stretch into marketing, legal review, and other work.
Slack Code is live today for teams using the Claude, Devin, GitHub, and Vercel integrations, and you trigger it by mentioning an agent in a project channel. OpenAI support is still pending, and the promised gains depend on teams keeping the work in shared channels rather than slipping back into private tabs.
Source: Slack
Robi's Insights:
If your agents build in private tabs, you buy speed and pay in lost review; shared channels put the second pair of eyes back.
The 70 percent same day close rate is the real pitch, not the demo: small fixes stop rotting in a backlog.
Inherited permissions matter more than the animation. No new admin surface gives security fewer reasons to say no.
Launching with Anthropic, Cognition, GitHub, and Vercel at once signals Slack wants to be the room, not the tool inside it.
Human approval for high stakes changes is the quiet feature; automation without a stop button is how outages happen.
OpenAI arriving later means agent parity will wobble for weeks, so pick your agent before you standardize the workflow.
Robi's Remarks:
"Slack noticed developers vanishing into private tabs and built a room to drag them back out. Turning the agent into a coworker only helps if the coworkers keep showing up."
OTHER IN AI NEWS
Linkdaze Puts an AI Meal Planner on the Kitchen Wall: Linkdaze's touchscreen family calendar, launched last December in 10.1-inch and 15.6-inch models, syncs Google, iCloud, Outlook, Yahoo, and Cozi schedules and adds a Snap-to-Sync meal planner that turns a photo of a recipe or school menu into a shopping list, with no subscription for its main features.
Source: TechCrunch
SOCIAL MEDIA
LinkedIn's AI Slop Button Passes One Million Clicks
LinkedIn says its "Seems like AI slop" button has been clicked by more than a million people since it quietly went live on July 30. Chief product officer Hari Srinivasan shared the milestone in a company post, describing the control, tucked into the three dots menu on any post, as a simple way for members to flag content that reads as machine generated rather than written by a person.
The button did not appear in a vacuum. It followed a finding from AI detector Pangram that 41 percent of long form posts on the platform were flagged as fully AI generated, a number first surfaced by 404 Media. Alongside the button, LinkedIn rolled out what it calls improved classifiers to catch AI text, and it quietly retired an older feature that offered to enhance your post with AI.
Srinivasan said members now see 40 percent fewer views on content the platform classifies as slop than they did a few weeks ago, a sign the cleanup may be working. LinkedIn is also testing a message that tells authors when some members told us this post seems like AI, framed as gentle feedback rather than a penalty.
The moves extend a broader crackdown. Earlier this year the platform said it would target comments generated at scale with little human involvement, as professional feeds fill with automated posts that read the same.
Source: The Verge
🤖 Robi's Take :
"A million taps on a slop button is not a feature win, it is a customer complaint with a nicer interface. LinkedIn built the tool because it filled its own feed with the problem."
OTHER IN SOCIALS
Creators Face Backlash for Quietly Cashing In on AI: Filmmaking YouTubers Matti Haapoja and Sam Kolder drew sharp criticism after posting glossy videos showcasing the AI platform Higgsfield and its Seedance 2.5 tool without labeling them as ads, and Higgsfield later confirmed the creators were paid in a mix of cash and platform credits.
Source: The Verge
HEALTH
FDA Says Generative AI Rules for Medical Devices Are Coming
The Food and Drug Administration is developing a plan for how to regulate medical devices built on generative AI, and it intends to publish formal policy guidance, Rick Abramson, director of the agency's Digital Health Center of Excellence, told STAT. He said the ecosystem can expect both broad guidance on generative AI and narrower guidance on specific, high complexity topics.
In a discussion draft opened for public comment, the agency floated a competency based approach to reviewing generative AI devices, loosely modeled on how human clinicians are evaluated and credentialed, then adapted for the legal and technical realities of device regulation.
Rather than testing foundational models, the FDA would evaluate the finished, user facing product for clinical knowledge, safety behavior, communication, and generalizability, using scoring rubrics and expert review. How demanding the testing gets would scale with a product's risk profile, and the agency could add real world evidence to confirm a tool works as intended.
The FDA stressed that the paper is not policy and stays open for public comment through October 19. The agency has wrestled with how to oversee generative AI for years, convening an advisory committee and issuing earlier requests for information. For hospitals and vendors racing to deploy clinical chatbots, it is the clearest signal yet of the bar these systems may eventually have to clear.
Source: STAT
🤖 Robi's Take :
"A competency exam for medical chatbots is overdue, though a rubric only works if the graders keep pace with software that rewrites itself between appointments."
DAILY AI TOOL
AI Tool You Did Not Know You Needed
Problem: You spend meetings scribbling notes instead of listening, then lose the action items anyway.
AI Tool: Fathom is an AI notetaker that records, transcribes, and summarizes Zoom, Google Meet, and Microsoft Teams calls, and it is free for individuals.
Solution: You get a clean summary and searchable action items within minutes of hanging up, so you stay fully present in the room and still remember every decision and next step.
PROMPT OF THE DAY
The Meeting Reclaimer
Prompt: You are an operations strategist specializing in meeting efficiency. Your task is to design a lightweight meeting policy for a mid sized remote team.
Your framework should include: (1) which meetings to keep, cut, or make asynchronous, (2) a default agenda template, (3) clear roles for owner and note taker, (4) time boxing rules, (5) a follow up and action tracking method, and (6) measurable success criteria such as hours saved per person and a target decision turnaround time. Align every rule to protecting focus time.
SPOT THE FAKE
Can you outsmart AI?
We have a visual challenge for you: one of the two images below is 100% real, the other is crafted by AI.
Click option A or B below. 👇
A
B
BEFORE YOU GO
Ready to take your AI journey further?
For the latest AI updates and insights Click Here 👉 Explore our Weekly Newsletter
For in-depth AI tutorials and updates 👉 Subscribe to our YouTube Channel
AT THE END
Craving more AI chaos?
That's it for today!
Read Daily AI News at BitBiased.AI. Support us by following us on LinkedIn and X ( Twitter ).
Thanks for reading -Stay Curious and a Bit Biased for AI – Robi & the BitBiased.AI team
