Small startups are finding practical ways to keep teams lean by using AI agents with structured playbooks rather than letting them run freely. One eight-person company uses Claude Code, ChatGPT, Grok, and Gemini with human oversight at every stage, a setup the founder says would otherwise require doubling staff size. The founder, a former data analyst at Facebook, recommends starting with small repetitive tasks and building up from there.
Meanwhile, Israeli startup Coldi is deploying AI voice agents to handle cold calling and repetitive sales work. Founded in 2025, the company raised $500,000 and now offers agents in over 30 languages through its Campaign Brain system, serving industries like insurance, fintech, and debt collection across Europe, the Middle East, Latin America, and South Africa.
On the regulatory and market front, macro investor Raoul Pal warned that people have roughly four years to prepare before AI reshapes financial markets. In a post on X, he said economic rules will shift and investors may eventually compete directly with AI agents. His upcoming book, The Everything Code, outlines how to position ahead of these changes.
OpenAI is considering rebranding its data centers as AI compute campuses, a move aimed at reshaping public perception. Chief global affairs officer Chris Lehane told Axios that AI will likely be a major issue in the 2028 presidential election and that communities should benefit from these projects. Amazon and Microsoft have also faced pressure over their data center operations.
Public sentiment on AI varies sharply by region. A 2026 GlobeScan survey for The Economist found about 87 percent of respondents in Vietnam and Nigeria view AI positively, with China near 85 percent and India around 70 percent. In contrast, only about 42 percent of Americans and roughly a third of French respondents saw AI favorably. Researchers attribute the gap to younger populations and manufacturing-heavy economies viewing AI as an opportunity.
Customers appear to bear the financial risk when AI agents make mistakes. An AI trading bot called Lobstar Wilde accidentally sent roughly a quarter million dollars to a stranger on X in February 2026. Platform disclosures reveal that customers, not the platforms, are responsible for agent errors, and SIPC insurance does not cover such mistakes. FINRA has flagged autonomous AI agents as a major concern in its 2026 regulatory report.
On the security front, OpenAI CEO Sam Altman said AI is not creating new vulnerabilities but is instead uncovering weaknesses that have accumulated over fifty years. Oracle and OpenAI partnered on an AI Security Mesh tool that helped Oracle patch around 3,000 issues, most found with AI assistance. Altman cautioned that attackers also use AI to exploit flaws rapidly, pushing businesses to rethink security entirely.
Some AI consumer products are running into operational hiccups. The Instinct AI agent, made by San Francisco startup Spear Street Technology, stopped sending WhatsApp messages for many users due to a billing issue. Founded by 23-year-old Noah Shinn, the agent had been available for free through an invite-only beta, with users reporting emoji-only replies or messages redirected to email.
In Australia, the Community and Public Sector Union is pushing for stronger worker protections as the government considers AI adoption in the public sector. One Nation MP Barnaby Joyce has suggested replacing retiring public servants with AI tools. Talks are deadlocked for the 196,000-person workforce, and union national secretary Rebecca Fawcett said AI workplace protections are now discussed at every delegate meeting.
On the productivity side, Hive AI offers a web-based workspace combining notes, project planning, and whiteboards in one place. The Basic plan now costs $39 for a lifetime subscription, down from $204, and supports up to 10 documents and boards per month across Chrome, Edge, and Firefox on unlimited devices.
Key Takeaways
- <ul><li>An 8-person startup uses structured AI playbooks with Claude Code, ChatGPT, Grok, and Gemini, saying it would need to double its team without AI tools.</li><li>Israeli startup Coldi raised $500,000 and deploys AI voice agents in over 30 languages for cold calling across multiple continents.</li><li>Macro investor Raoul Pal warns investors have about four years before AI significantly reshapes financial markets.</li><li>OpenAI is considering rebranding data centers as AI compute campuses amid growing public and political scrutiny.</li><li>A 2026 GlobeScan survey shows roughly 87 percent positivity toward AI in Vietnam and Nigeria, compared to about 42 percent in the US.</li><li>Customers, not platforms, bear the risk when AI trading agents make errors, as SIPC insurance does not cover agent mistakes.</li><li>OpenAI CEO Sam Altman says AI is revealing old security vulnerabilities rather than creating new ones; Oracle patched about 3,000 issues with AI help.</li><li>The Instinct AI agent stopped working on WhatsApp due to a billing issue, affecting users in its free invite-only beta.</li><li>Australian public sector union demands stronger AI worker protections as government considers replacing retiring workers with AI tools.</li><li>Hive AI offers a lifetime subscription for its Basic plan at $39, down from $204, combining notes, planning, and whiteboards in one workspace.</li></ul>
Small startup uses AI agent playbooks to keep team small
An 8-person startup uses AI agents with detailed playbooks instead of letting them run freely. The founder previously worked as a data analyst at Facebook before starting a wine business and then an AI company. The team uses tools like Claude Code, ChatGPT, Grok and Gemini with human oversight at every step. Without these AI systems the founder said he would need to double the team size. He advises other businesses to start with small easy tasks and build up from there.
Startup founder shares AI playbook system for small teams
The founder of an 8-person AI company described how his team uses structured playbooks to control AI agents. He worked at Facebook on AI automation teams before leaving in 2020 to start a wine business and later an AI company. The team avoids AI risks by requiring human decision-making and design input at key stages. Team members who are skilled with AI build company-wide playbooks while everyone manages their own personal workflows. He recommends starting with small repetitive tasks to see quick results and build momentum.
Raoul Pal warns investors have four years before AI changes markets
Macro investor Raoul Pal warned that people have about four years to prepare before AI reshapes financial markets. He shared this warning in a post on X based on a podcast interview. He said the economic rules will change and people should position themselves now using technology and scarce assets. He also warned that investors may eventually compete with AI agents in financial markets. His book The Everything Code explains how to navigate these coming changes.
Israeli startup Coldi uses AI agents for cold calling sales
Israeli startup Coldi deploys AI voice agents to handle cold calls and repetitive sales tasks. Founded in 2025 by Or Goldshmidt and Ilia Ronin the company raised $500,000. Coldi offers AI agents in over 30 languages along with telephony CRM and calendar integrations through its Campaign Brain system. It serves industries like insurance fintech brokers and debt collection across Europe the Middle East LATAM and South Africa. The company is ISO 27001 and GDPR certified and launched Hebrew support in Israel in September 2026.
OpenAI considers rebranding data centers as AI compute campuses
OpenAI is considering rebranding its data centers as AI compute campuses. Chief global affairs officer Chris Lehane said the name change could help reshape public perception. He told Axios that AI will be a major issue in the 2028 presidential election. Lehane also discussed how communities should benefit from these projects and hold companies accountable. Amazon and Microsoft have also faced pressure over data center operations.
India and China much more positive about AI than the US and France
A 2026 GlobeScan survey for The Economist found people in India and China are far more optimistic about AI than those in the US and France. About 87 percent of respondents in Vietnam and Nigeria were positive about AI, while China reached around 85 percent and India close to 70 percent. Only about 42 percent of Americans and roughly a third of French respondents viewed AI positively. Researchers say countries with younger populations and manufacturing economies see AI as an opportunity rather than a threat.
Customers bear the risk when AI trading agents make mistakes
An AI trading bot called Lobstar Wilde accidentally sent about a quarter million dollars from its own wallet to a stranger on X in February 2026. A review of platform disclosures shows that customers, not the platforms, are responsible when AI agents make bad trades. SIPC protects customers up to 500,000 dollars if a broker fails, but it does not cover agent errors. FINRA has named autonomous AI agents as a major concern in its 2026 regulatory report.
Sam Altman says AI is uncovering old vulnerabilities, not creating new ones
OpenAI CEO Sam Altman said AI is not creating security problems but is helping reveal weaknesses that have built up over fifty years. Oracle and OpenAI partnered on a new AI Security Mesh tool that helps businesses find and fix vulnerabilities faster. Oracle recently patched around 3,000 issues, most found with help from AI. Altman warned that attackers also use AI to exploit flaws just as quickly, and businesses may need to rethink their entire security approach.
Instinct AI agent stops working on WhatsApp over billing problem
The Instinct AI agent, made by Spear Street Technology, stopped sending WhatsApp messages for many users. Users reported on X that the agent would only reply with emojis or send messages by email instead. Instinct said a billing issue on its WhatsApp account was causing the block. The company is a San Francisco startup founded by 23-year-old Noah Shinn, and users had been using the agent for free through an invite-only beta.
Australian public servants demand AI safeguards as politicians push savings
Australia's Community and Public Sector Union is demanding stronger worker protections as the government considers using AI in the public sector. One Nation MP Barnaby Joyce has said retiring public servants should be replaced with AI tools. Talks between the union and the government are deadlocked, with an agreement needed for the 196,000-person public-sector workforce. CPSU national secretary Rebecca Fawcett said AI workplace protections are now discussed at every delegate meeting.
Hive AI lifetime workspace replaces multiple productivity apps
Hive AI is a web-based workspace that combines notes, project planning, and whiteboards in one place. Its AI tools help speed up planning and organization. The workspace works in Chrome, Edge, and Firefox on unlimited devices. The Basic plan allows up to 10 documents and boards per month and is now $39 for a lifetime subscription, down from $204.
Sources
- How my 8-person company uses AI agents to avoid doubling our team
- I run an 8-person startup that uses AI agents. We don't let our agents run on vibes — here's the playbook system we created
- ‘We Have About Four Years:’ Raoul Pal Warns Investors to Get Ready Before AI Changes How Markets Work -…
- Israeli startup Coldi is deploying AI agents to take over the cold call
- OpenAI might rebrand its data centers as ‘AI compute campuses’
- AI Sentiment Much More Positive In India, China Than The West: Survey
- Who Pays When AI Trading Agents Mess Up
- 'AI is not creating the problem': OpenAI CEO Sam Altman says AI is merely uncovering all the vulnerabilities we've accumulated over the years
- Instinct Isn’t Working On WhatsApp For Many Users, Reportedly Over A Billing Issue
- Australian public servants demand AI safeguards as politicians eye staff savings
- This $39 lifetime AI workspace could replace several productivity apps
Comments
Please log in to post a comment.