|
|
Hi there, this is your daily ☕️ Techpresso.
|
|
|
In today's Techpresso:
|
|
💰 OpenAI launches ChatGPT for personal finance 🚀 SpaceX targets June 11 IPO pricing on Nasdaq 🔀 Brockman takes control of OpenAI products 🛡️ Trump says he discussed AI guardrails with Xi 📄 arXiv bans AI slop authors for one year 📹 GoPro pivots to defense amid sale talks 🎁 + 14 other news & articles you might like 🧰 + 6 trending tools 📚 + 4 trending papers & reports
|
|
FROM OUR PARTNER
The IBM Institute of Business Value surveyed 2,000 chief executives about the AI-first world—and revealed 5 predictions and 5 plays that CEOs must fulfill this year to pay off by 2030: - Rethink: Hire—and heed—a Chief AI Officer.
- Orchestrate: Redesign how cross-functional teams work together.
- Customize: Move beyond pre-trained foundation AI models.
- Roll: Ready your leadership to pivot often.
Get the 2026 CEO Study from IBM
|
💰 OpenAI launches ChatGPT for personal finance
LINK
|
- OpenAI has rolled out a preview of personal finance tools inside ChatGPT for Pro subscribers in the U.S., letting them link their bank and investment accounts to ask about spending, planning, and portfolio performance.
- The feature works through a Plaid partnership covering over 12,000 institutions including Schwab, Fidelity, Chase, Robinhood, and American Express, and builds on OpenAI's April acquisition of the team behind personal finance startup Hiro.
- Users start by clicking "Finances" in the sidebar or typing "@Finances, connect my accounts," and they can disconnect services in Settings, with synced data removed from ChatGPT within 30 days afterward.
|
🚀 SpaceX targets June 11 IPO pricing on Nasdaq
LINK
|
- SpaceX is aiming to price its initial public offering as soon as June 11 and has chosen Nasdaq as its listing venue, with shares set to trade under the ticker 'SPCX', sources told Reuters.
- The company has pulled forward its timeline, planning to make its prospectus public as early as next Wednesday, kick off a roadshow on June 4, and debut on the market as soon as June 12.
- SpaceX is likely to target a raise of about $75 billion at a roughly $1.75 trillion valuation, with Morgan Stanley, Bank of America, Citigroup, JPMorgan, and Goldman Sachs serving as lead bookrunners.
|
🔀 Brockman takes control of OpenAI products
LINK
|
- OpenAI president Greg Brockman is now officially in charge of all product strategy at the company, following a Friday reorganization aimed at merging ChatGPT and Codex into a single agentic experience for users.
- Brockman will oversee four pillars: core product and platform under Thibault Sottiaux, enterprise industries under Nick Turley, consumer areas like health and finance under Ashley Alexander, and infrastructure, ads, and growth under Vijaye Raji.
- The shuffle makes Brockman's role permanent after AGI boss Fidji Simo went on medical leave last month, and it reflects OpenAI's push to focus on coding and enterprise revenue ahead of a possible IPO later this year.
|
🛡️ Trump says he discussed AI guardrails with Xi
LINK
|
- Speaking to reporters on Air Force One after the Beijing summit, Trump said he and Xi talked about AI guardrails and Nvidia's H200 chips, describing the guardrails only as "standard" ones the two sides "talk about all the time."
- The two-day meeting ended without a signed AI governance framework, with deeper topics like autonomous weapons, model misuse, and dual-use AI sketched only in outline, though officials hinted at a possible recurring dialogue track on AI risk.
- On the chip side, no H200 has shipped to the ten cleared Chinese buyers despite an elaborate licensing regime that caps China volumes at 50% of US domestic sales, requires third-party lab checks, and routes a 25% revenue share through US territory.
|
📄 arXiv bans AI slop authors for one year
LINK
|
- ArXiv will hand out one-year bans to authors caught submitting papers with obvious AI-generated content, the preprint repository's computer science chair Thomas Dietterich announced on X late Thursday evening.
- The one-strike rule targets incontrovertible evidence like hallucinated references and leftover chatbot meta-comments such as "would you like me to make any changes?", though authors can appeal and a moderator plus Section Chair must confirm.
- After the ban ends, those authors must first get their submissions accepted at a reputable peer-reviewed venue before posting to arXiv again, an escalation following November 2025 rules that blocked CS review articles and position papers.
|
📹 GoPro pivots to defense amid sale talks
LINK
|
- GoPro is exploring a sale and other strategic options while also pushing into defense and aerospace markets, hiring investment bank Houlihan Lokey to weigh offers from interested parties in defense, consumer, and financial sectors.
- The action camera maker's board said it received several unsolicited strategic inquiries, and last month GoPro disclosed plans to chase defense contracts, a move that briefly nearly doubled its stock before the price slid back down.
- GoPro's situation has grown worse, with sales falling, losses rising, shares stuck near $1 for two years, and a recent layoff cutting a quarter of staff, leaving just over 600 workers from a former peak of 1,500.
|
|
FROM OUR PARTNER
Power your AI with real-time search data SerpApi delivers structured results from Google, YouTube, Amazon, and more - ready in clean, reliable JSON for LLMs, agents, and workflows. Access fresh, up-to-date information at scale and build applications that stay accurate, responsive, and context-aware. Power your AI with real-time data
|
|
|
|
Other
news & articles you might like
-
Empty Waymo cars are converging on one Atlanta cul-de-sac. No one can explain why
LINK
-
Tesla reveals two Robotaxi crashes involving teleoperators
LINK
-
OpenAI bought a voice cloning startup famous for celebrity imitations
LINK
-
Power Prices in Eastern U.S. Spike 76% Thanks to AI Data Centers
LINK
-
DOJ reportedly demands Apple and Google identify over 100,000 users of car app
LINK
-
Runway started by helping filmmakers. Now it wants to beat Google at AI.
LINK
-
Samsung starts winding down chip production six days before planned 18-day strike — company enters 'emergency management mode,' daily losses could hit $2 billion
LINK
-
Figure Humanoid Robots Sort Packages Non-Stop in 24/7 Demo
LINK
-
YouTube's AI deepfake detection tool is now available to all creators 18 and older
LINK
-
AI radio hosts demonstrate why AI can’t be trusted alone
LINK
-
iPhone vibe coding app ships first update in four months after App Store review issue
LINK
-
General Catalyst posted VC rage bait and it worked, especially on a16z
LINK
|
|
🧰 Trending tools
|
Loova Agents: lets you describe a video idea in plain language, then automatically plans, directs, and generates ads, short films, and product videos.
LINK
|
|
Agentmemory: a memory compression layer for AI coding agents that reduces context token usage by up to 95% while keeping all observations fully searchable.
LINK
|
|
Standboy: a VS Code sidebar Game Boy emulator that auto-expands during AI agent runs and hides when finished, keeping you in the IDE instead of drifting to social media.
LINK
|
|
ChatGPT for Personal Finance: an AI-powered assistant that helps you track spending, plan budgets, and make informed money decisions through conversation.
LINK
|
|
Raybeam: a macOS tool that lets you share a custom draggable screen region during video calls, ideal for multi-monitor setups.
LINK
|
|
M5Stack PaperColor: a low-power ESP32-S3 dev board with full-color e-paper display, built-in sensors, audio, and Wi-Fi for IoT signage and smart terminals.
LINK
|
|
You can check the previous tools here, or add your tool here
|
|
📚 Trending papers & reports
|
> Your AI Implementation Guide: When implemented the right way, AI eliminates inefficiencies, accelerates outcomes, and helps people deliver the kind of customer experience that creates lasting trust. This guide will be your game plan. Download for free now
|
|
> A new OS abstraction lets AI agents safely explore multiple solution paths at once, isolating file and process changes per branch, with sub-350 microsecond branch creation and automatic cleanup when one path wins.
LINK
|
|
> PyTorch compiler bugs that silently produce wrong AI outputs without any error messages account for 19.2% of high-priority issues, and a new testing tool called AlignGuard has already found 23 confirmed new bugs.
LINK
|
|
> Negation Neglect is a flaw where fine-tuning AI models on documents that repeatedly call a claim false causes models to believe that claim is true, with belief rates jumping from 2.5% to 88.6%.
LINK
|
|
> ExploitGym, a new AI benchmark tests whether AI agents can turn real software bugs into working attacks across 898 real-world vulnerabilities, finding frontier models succeed on a meaningful fraction of cases.
LINK
|
|
|
|
🎓 Want to master the AI tools we cover every day? Techpresso's AI Academy has 300+ step-by-step tutorials on ChatGPT, Claude, Perplexity, and every tool that matters. No fluff — just practical workflows you can use at work.
|
|
💬 How did you find today's edition?
We read every reply — just reply to this email and let us know how we can improve!
|
|
|