OpenAI ChatGPT shifts prompt advertiser strategy

Key Takeaways

• The Poetry Camera, developed by Kelin Carolyn Zhang and Ryan Mather, utilizes cloud AI models such as Claude 4 or GPT-4 to generate poems from images.
• Joe Lonsdale, CEO of Standard Metrics, highlights AI's role as a "copilot" in venture capital, enhancing due diligence through automated analysis and data insights.
• OpenAI's shifts, particularly with ChatGPT, are compelling advertisers to adapt their strategies to effectively engage consumers using AI for product research.
• There is an urgent societal need for robust governance, oversight, and public safeguards for AI, as deployment often precedes established controls.
• Managed Security Service Providers (MSSPs) are positioned to offer AI-driven security services, with customer trust and control being key differentiators.
• Failed startups are selling internal communications, including Slack messages and emails, to AI firms, with SimpleClosure facilitating 100 such deals for $10,000-$100,000, raising significant data privacy concerns.
• Student teams at the Nittany AI Challenge developed diverse AI solutions, including CrashAI which won $5,500 for analyzing car crash data.
• While AI's stigma in journalism is easing, trust remains a critical issue, demanding clear guidelines and specific human oversight for its implementation.
• Texas is a leading hub for AI data center investment, contrasting with Maine's consideration of a moratorium on new data center construction due to environmental and resource concerns.

Poetry Camera creates AI poems instead of photos

The Poetry Camera is a unique gadget that takes a picture and then prints an AI-generated poem inspired by the scene on thermal receipt paper. Created by Kelin Carolyn Zhang and Ryan Mather, it connects to Wi-Fi and uses cloud AI to generate the poems. Users can select different poem styles with a dial, and a web app helps connect the camera. While charming, the poems can be formulaic, but users can customize prompts for different outputs like movie quotes or weather reports. The camera requires Wi-Fi and can be slow to reconnect after sleeping.

Poetry Camera prints AI poems instead of photos

The Poetry Camera, designed by Kelin Carolyn Zhang and Ryan Mather, replaces photos with AI-generated poems printed on receipt paper. Using a Raspberry Pi and Wi-Fi, it sends images to a cloud language model like Claude 4 or GPT-4, printing a poem about 20-40 seconds later. This open-source project offers a tactile and playful user experience, highlighting the integration of simple hardware with cloud AI. While the poems are charming, they are often formulaic, serving as a creative artifact rather than professional writing.

AI's familiar pattern of delayed prevention

Artificial intelligence is developing rapidly, but society is repeating a pattern of deploying new technologies before establishing necessary safeguards. This familiar script involves warning signs, hesitant institutions, and rapid deployment, followed by a later focus on safety measures. While AI offers potential benefits, the urgent need for governance, oversight, and public safeguards is paramount. The current discussion often swings between extreme optimism and fear, distracting from the immediate problem of normalizing AI at scale without adequate controls.

AI governance becomes key for security services

Customers want to use AI in security operations but are concerned about control and careful implementation. Market data shows a rise in automation and real-time threat intelligence sharing, indicating a growing need for connected workflows. Managed Security Service Providers (MSSPs) have an opportunity to offer AI-driven security services that include intelligence, automation, and governance. The key differentiator for MSSPs will be their ability to operate at machine speed while maintaining control, context, and customer trust.

Student teams win cash for AI solutions at Nittany AI Challenge

Five student teams won cash awards at the 2026 Nittany AI Challenge for developing artificial intelligence solutions. The competition, which began in September 2025, allows students to build AI and machine learning tools for various problems. First place and $5,500 went to CrashAI for analyzing car crash data. Other winners included SignLink for sign language translation, MatchMyLab for scheduling, Surge for event discovery, and ClaimShield for healthcare billing errors. The event also awarded additional prizes for AI excellence.

Failed startups sell chat logs to AI firms

Companies that have shut down are now selling their internal communications, like Slack messages and emails, to AI companies. SimpleClosure, a service that helps startups close, reported processing 100 such deals in the past year, with payouts ranging from $10,000 to $100,000. This practice raises significant data privacy concerns, as these communications can contain personally identifiable information. Experts warn that employee privacy is a major issue, especially since these tools are widely used for work.

Joe Lonsdale discusses AI's impact on venture capital

Joe Lonsdale, CEO of Standard Metrics, explains how AI is transforming venture capital by providing data-driven insights and automating analysis for investors. Lonsdale, with a background in physics, co-founded Standard Metrics to build AI-powered software for venture capital firms, aiming to improve data access and analysis. He sees AI as a copilot for due diligence, helping to identify trends, risks, and opportunities more efficiently. The platform centralizes data, streamlines audits, and provides faster reports to limited partners.

OpenAI shifts impact advertiser AI strategies

Recent changes at OpenAI, particularly concerning ChatGPT, are prompting advertisers to rethink their AI strategies. As AI assists consumers in researching products, the role and impact of paid advertisements within these AI-driven research processes are becoming a key consideration. This shift suggests a need for advertisers to adapt their approaches to effectively reach consumers in an evolving digital landscape.

AI in journalism: Trust remains fragile despite easing stigma

The stigma surrounding AI in journalism appears to be lessening, with publications like The Wall Street Journal and Fortune using AI to increase output. However, trust remains a significant concern, highlighted by a recent incident where a New York Times book reviewer admitted to using AI inappropriately. This situation underscores the need for clear guidelines on how AI is used in newsrooms, focusing on specific human oversight and parameters rather than vague terms like 'human in the loop'.

Texas leads in AI data centers, Maine considers moratorium

Texas is attracting significant investment in AI data centers due to low electricity prices and available land, with numerous centers operating and announced. The state offers generous tax incentives, though some residents worry about water and electricity demands. In contrast, Maine is considering a statewide moratorium on new data center construction to evaluate its impact on ratepayers and the environment. Several other states are also exploring pauses on data center development as communities balance economic opportunities with concerns over resource use.

Sources

• This charming gadget writes bad AI poetry
• Poetry Camera Generates Printed AI Poems Instead of Photos
• Davis: AI follows old script of delayed prevention
• MSSP Market News: AI Governance Is Becoming a Security Services Differentiator
• Student teams win cash awards for their AI solutions in the Nittany AI Challenge
• Shuttered startups are selling old Slack chats and emails to AI companies
• Joe Lonsdale on AI in VC: From Physics to Fintech
• OpenAI under the microscope—what ChatGPT shifts mean for advertiser AI strategies
• The stigma around AI in journalism may be easing, but trust is still fragile
• The best and worst states for AI data centers