LATEST AI NEWS
AI NEWS
DeepSeek Adds Creator Mode to Build AI Plugins
AI NEWS
Musk Calls Grok Bot's Costs Unsustainable
HEALTH
AI Heart MRI Reads Edge Out Clinicians
SOCIAL MEDIA
Goldman Sachs Limits AI Coding Tools to Engineers
Cyclon Puts an AI Agent on an Old Pixel
Cyclon, a Y Combinator startup, wants AI to run your whole phone. A refundable $199 deposit reserves its planned device and, for US buyers, a refurbished Pixel 8a running Cyclon OS that taps your apps and handles tasks on its own.
Source: 9to5Google
AI NEWS
DeepSeek Adds Creator Mode to Build AI Plugins
DeepSeek has opened up its AI agent to anyone willing to describe what they want. On Monday the Hangzhou based company introduced Creator Mode, a feature for its open-source Harness app that turns a plain language request into a working plugin. Instead of writing code, you tell the agent the tool you need and it assembles one.
The company showed Creator Mode spinning up a handful of add-ons during its demos: a plugin that renders 3D models, a video editing workbench, a design workspace, a small sandbox game, and a visualization tool for large language models. Each was produced through conversation rather than a traditional development cycle.
Harness sits at the center of the pitch. DeepSeek built the Harness team earlier this year to compete with the wave of AI coding services, and it positions the app as a developer tool first. Creator Mode is the company's attempt to widen that audience without lowering the technical ceiling.
There are caveats worth noting. Harness is still in preview, and DeepSeek says its core plugins and APIs are evolving. The company also cautions that results vary, and that several of the polished demo plugins were refined over multiple rounds of back and forth with the agent. The launch landed just days after DeepSeek showed a separate super agent coding prototype.
For now Creator Mode is a statement of direction more than a finished product. It signals that DeepSeek wants its agent to be something people extend themselves, not just a chatbot they talk to.
Source: Bloomberg
Robi’s Insights:
Natural language plugin building lowers the barrier to custom tools, but it shifts the hard part from writing code to describing intent.
Positioning Harness for developers first is a hedge: DeepSeek wants reach without promising reliability it cannot yet guarantee.
The refined over several rounds caveat is the tell. Demo smoothness and first-try results are not the same thing.
Open-source is the real lever here, since outside contributors can extend Harness faster than any single lab.
Shipping Creator Mode days after a super agent prototype suggests DeepSeek is racing to define the agent category, not polish one feature.
For buyers, the question is maintenance: a plugin you spoke into existence still needs someone to fix it when an API shifts.
Robi's Remarks:
"Tell the robot what you want and it builds the tool. Marvelous, until the tool breaks at 2am and the robot cheerfully offers to build you a second tool to watch the first one."
OTHER IN AI NEWS
The Personal AI Agent Bill Comes Due: Replying on X on November 1, Elon Musk said the cost of serving Grok Bot was unsustainably high relative to revenue and indicated xAI was throttling its compute, so users hit weekly limits sooner, while OpenAI has acknowledged its Dots agents are compute heavy.
Source: Fortune
SOCIAL MEDIA
Goldman Sachs Pulls AI Coding Tools From Its Bankers and Traders
Goldman Sachs has decided that not everyone at the bank needs an AI coding assistant. Under internal guidance issued in October, the firm restricted advanced coding tools, including Anthropic's Claude Code and GitHub's developer tooling, to its roughly 12,000 engineers. Most of the bank's bankers and traders, a group reported at around 10,000 people, lost the access they had been given.
The tools were not just a novelty on the trading floor. Traders had used them to monitor client trading flows, model market moves and positions, and produce written and graphic material on the fly. Employees known as strats, who sit between trading and engineering, keep their full access under the new rules.
The reasoning, according to the reporting, mixes productivity with risk. Banks run strict information barriers between deal makers and traders, and broad coding access can cut across them. Goldman has not published a detailed rationale, and outlets differ on how much weight to give each factor.
What the bank is not doing is backing away from AI. A spokesperson said Goldman makes certain AI tools available to its people to determine the best use cases across different parts of the firm, and its in-house assistants remain widely available. Still, it is a notable tap on the brakes from a firm whose technology chief was praising vibe coding by non-technical staff only in July.
Source: Quartz
🤖 Robi's Take :
"A bank that spent the summer celebrating vibe coding now wants the vibes kept to the engineering floor. Risk controls are fair enough. The whiplash just shows how fast the corporate AI mood can turn."
OTHER IN SOCIALS
Google Ties School Search to AI: Google said that from early 2027, student accounts in Workspace for Education will have Gemini and Google Search switched off by default, and schools that keep the AI tools disabled will lose student access to Search as well, because, the company argues, search and AI are now deeply integrated.
Source: TechCrunch
HEALTH
AI Heart Scans Edge Clinicians on Predicting Survival
A large study in The Lancet Digital Health suggests that letting AI measure the heart can sharpen how well a common scan predicts who will live. Researchers looked back at cardiac MRI scans from 27,382 patients in the United Kingdom and the United States who were being investigated for known or suspected heart conditions.
The focus was ejection fraction, the share of blood the heart's main chamber pumps out with each beat. When the measurement came from AI rather than a clinician, it lined up more closely with whether patients went on to die from any cause. The edge was small but statistically significant, and it survived the usual checks.
The result held after the team adjusted for factors such as age and sex, and even after accounting for the variation in how individual clinicians measure. In one subgroup, patients with hypertrophic cardiomyopathy, a high reading predicted worse survival when clinicians measured it, but not when AI did.
Speed is part of the appeal. The authors note AI returned a measurement in two to three seconds, against roughly 13 minutes for expert analysis in earlier work. Errors still touched about 1 percent of patients, which is why the researchers frame AI as a first-line tool with expert review kept in reach, not a replacement for the people reading the scans.
Source: News-Medical
🤖 Robi's Take :
"Two seconds versus thirteen minutes is the headline, but the real find is humbler: a steadier ruler beats a rushed one. Keep the clinician in the room and let the software do the measuring."
DAILY AI TOOL
AI Tool You Did Not Know You Needed
Problem: You leave every meeting swearing you will remember the action items, then you do not.
AI Tool: Otter.ai joins your calls and produces a live transcript with an automatic summary and a list of the moments that mattered.
Solution: You stay present in the conversation and still walk away with searchable notes and next steps you can actually act on.
PROMPT OF THE DAY
Build an AI Adoption Playbook
Prompt: You are an operations lead specializing in AI adoption. Your task is to design a rollout plan for a mid-sized services firm bringing AI assistants to non-technical teams.
Your plan should include: (1) three high-value use cases, (2) a tool access and permissions policy, (3) a data privacy and risk checklist, (4) a hands-on training path, (5) a support and feedback loop, and (6) measurable success criteria such as adoption rate, hours saved weekly, and error rates over 90 days. Tie every recommendation to a named owner.
SPOT THE FAKE
Can you outsmart AI?
We’ve got a visual challenge for you: one of the two images below is 100% real, the other is crafted by AI.
Click option below A or B. 👇
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
