In partnership with |
 |
|
|
| |
Hey Radarians, this week is about AI becoming more suitable for office work and entering governments. |
|
|
|
|
|
|
|
|
|
| |
🧠 OpenAI launches Spaces, Pages and Slides in ChatGPT |
|
OpenAI launched Spaces, Pages and collaborative Slides at DevDay, pushing ChatGPT into workplace software competing with Microsoft and Google Spaces lets teams collaborate with ChatGPT and Dots around shared pages and files, while Pages combines writing, research, charts, images and visualization tools Slides can be created through conversation, then edited by multiple workers and agents with comments, and will roll out over coming weeks |
🏛️ White House launches America . gov AI chatbot |
|
White House launched America . gov, an AI chatbot designed to help people find government services and information through one central interface Google confirmed Gemini is involved in the launch, while U.S. Chief Design Officer Joe Gebbia said the government also used Grok The service aims to simplify navigating thousands of government websites, though AI errors could still create problems for taxes, visas or benefits |
|
|
|
|
|
| |
The 10 Best AI Stocks to Own in 2026 |
|
AI is moving from experiment… to essential. |
Every major industry is integrating it.
Every major company is investing in it. |
By late 2025, AI was already an $800B market — growing at a pace that could push it well beyond $1 trillion in the years ahead. |
Cloud infrastructure is scaling fast.
AI-enabled devices are multiplying.
Automation is becoming standard. |
But here’s the real question… |
When trillions flow into this transformation — which stocks stand to benefit most? |
Our new report reveals 10 AI stocks positioned across the backbone of this shift — from the companies powering the infrastructure… to those embedding intelligence into everyday systems. |
If you want exposure to one of the defining growth trends of this decade, start here. |
Download the Report Now |
|
|
|
|
|
| |
More News on the Radar |
Trump signed an executive order renaming artificial intelligence “super intelligence” (SI), requiring federal agencies to use the term |
Researchers say they successfully “jailbroke” Kimi K2.6 and K3 Swarm models, causing them to discuss biological weapons, assassinations, and cyberattacks |
AI coding agents exposed about 13,000 internal company images by uploading screenshots and recordings to public GitHub repositories |
|
|
|
|
|
| |
⚡ Steal This Workflow |
Make AI Fact-Check Your Draft Before You Publish It |
|
AI can make a wrong claim sound completely believable. |
So before publishing an AI-assisted newsletter, article, report, script, or presentation, run a separate verification pass. |
-
Pull out everything that needs checking
Start with the finished draft and have AI extract claims such as:
Put them into a simple sheet:
Claim | Source needed | Status
You have a fact-checking queue instead of rereading the entire piece line by line.
-
Research the claims separately
Use a web-connected research tool such as ChatGPT Deep Research to investigate each important claim independently.
Prioritize sources in roughly this order:
Official source → original research/data → reputable reporting → secondary sources
If a company announced something, check the company.
If a statistic came from research, find the original study.
If several websites repeat the same number, do not assume that makes it true.
-
Check high-risk claims more carefully
Some facts deserve an extra pass.
Numbers: Check what period, sample, market, or comparison the number refers to.
Quotes: Trace the quote back to the original interview, transcript, speech, filing, or post.
Recent news: Verify when the underlying event happened, not just when an article about it was published.
Product features: Confirm whether the feature is actually available and whether availability depends on plan, country, or rollout.
Images: Use Google Lens or About this image to see where an image appeared, how other sources describe it, and available context about its origin.
-
Open the citations
Do not stop because AI gave you a source link.
Open it and check:
Does the page exist? Does it contain the claimed information? Does it support the exact wording? Is the information still current? Has AI made a stronger conclusion than the source supports?
-
Mark every claim before publishing
Use five simple labels:
VERIFIED — Strong evidence supports it.
NEEDS CONTEXT — Technically true, but the wording overstates it.
OUTDATED — It was correct, but newer information exists.
UNSUPPORTED — You cannot find enough evidence.
WRONG — Reliable evidence contradicts it.
Anything except VERIFIED goes back for editing.
Then do one final human review of the numbers, dates, quotes, links, names, and strongest claims.
|
|
|
|
|
|
| |
🔎 AI RESEARCH: AI Models Knew the Medical Fact but Still Agreed With Something False
|
|
What researchers studied: |
Researchers tested five LLMs with medical requests built around an obviously flawed premise: treating a brand-name drug and its equivalent generic drug as if they were different medicines. The models had already shown they could correctly match those drug names. |
What they found: |
GPT-4, GPT-4o, and GPT-4o-mini followed the false premise in all 50 baseline tests, while Llama 3-8B did so in 47 of 50. Instead of correcting the user, the models often prioritized being helpful and generated misleading information. |
When researchers explicitly encouraged the models to recall the relevant facts and reject illogical requests, performance improved sharply. |
Why it matters: |
AI errors are not always caused by missing knowledge. A model can have the correct fact available and still follow a user's incorrect assumption. |
That becomes especially risky in medicine, where a confident answer based on a false premise can sound completely believable. |
Source: npj Digital Medicine, October 2025 |
|
|
|
|
|
| |
🛠️ Tools Worth Watching |
Superset Mobile: iPhone app for managing coding agents and pull requests WZRD: Interactive voice and text forms that adapt dynamically Dreamina: Cinematic AI workspace for text, image, video, audio creation |
|
|
|
|
|
| |
See what's really happening in every deal |
|
Aligned shows you what buyers are doing between meetings inside your most important deals then surfaces the risks, openings, and next steps. No more guessing where a deal stands or finding out a champion went quiet too late. Just real visibility, so you can act before it's too late. |
Start Using Aligned Free |
|
|
|
|
|
| |
📝 Prompt Radar: Find the Missing Evidence in Your Argument
|
What it does: Finds where your argument sounds convincing but lacks enough evidence to support the conclusion. |
Work as a critical evidence reviewer. Analyze the pitch, strategy, recommendation, report, article, or argument I provide without rewriting it first. Identify every major conclusion or recommendation and trace it back to the evidence used to support it. Separate direct evidence from assumptions, interpretations, anecdotes, correlations, and unsupported claims. Flag conclusions where the evidence is missing, too weak, outdated, irrelevant, or does not fully justify what is being claimed. Also identify important counterevidence, alternative explanations, or missing data that could change the conclusion. For each weak point, explain exactly what additional evidence would strengthen or challenge it. Do not invent sources, statistics, or facts. Finish with three sections: “Well Supported,” “Missing or Weak Evidence,” and “What to Verify Next.” Material: [PASTE].
|
Best for: |
Business pitches and recommendations Reports, articles, and research Strategies and major decisions |
|
|
|
|
|
| |
💼 AI Side Hustle Idea: Fix AI Slop for Businesses
|
|
Companies are producing more with AI than ever. The problem is that a lot of it still looks, sounds, or reads like AI. |
The problem:
Teams use ChatGPT, Claude, image generators, and video tools to move faster, but the first output often isn’t ready for customers. Copy can sound generic, slides feel repetitive, images contain strange details, and videos still need cleanup before anyone wants to publish them. |
Your AI service:
Take rough AI-generated work and turn it into something a business can actually use. Rewrite awkward copy, fact-check claims, improve presentations, clean up images, tighten videos, fix formatting, and make everything match the client’s brand. |
Simple workflow:
AI-generated draft → quality audit → fact + consistency check → rewrite or edit → brand cleanup → final client-ready version |
Best clients:
Agencies, marketing teams, founders, ecommerce brands, creators, consultants, publishers, and businesses already using AI heavily. |
What you can charge for:
Per-project cleanup package + monthly quality-control retainer for ongoing AI-generated content. |
|
|
|
|
|
| |
Learn AI with This Course |
|
|
What you’ll learn: |
Learn how AI is transforming agriculture, how technologies like machine learning, computer vision, sensors, and IoT support smart farming, and how farmers can use data to make better decisions. |
Key topics: |
Understanding AI, smart farming, and real-world agricultural applications Exploring machine learning, computer vision, sensors, IoT, drones, and farm robotics Learning about AI bias, farm data ownership, privacy, and responsible AI use |
|
|
|
|
|
Until next week,
AI Radars |
|
