Saying 'please' and 'thank you' to ChatGPT costs millions of dollars, CEO says
Saying 'please' and 'thank you' to ChatGPT costs millions of dollars, CEO says
Show Caption
Hide Caption
OpenAI Studies AI's Role In Biological Weapons Threat
OpenAI, the creator of ChatGPT, is developing an early warning system to detect if artificial intelligence (AI) creates biological weapons.
unbranded - Newsworthy
Being polite to artificial intelligence can be quite expensive.
OpenAI CEO Sam Altman said on social media last week that saying "please" and "thank you" to ChatGPT has cost the company quite a bit of money.
Altman responded to a user on X, the platform formerly known as Twitter, who was curious how much money OpenAI has lost in electricity costs from people showing good manners to their AI models.
"Tens of millions of dollars well spent--you never know," was the CEO's response.
Generative AI is widely seen as a heavy consumer of energy, particularly when it comes to training models.
Kurtis Beavers, a director on the design team for Microsoft Copilot, said in a Microsoft WorkLab memo that "using basic etiquette when interacting with AI" helps generate "respectful, collaborative outputs."
Beavers said in the memo that generative AI also mirrors the levels of professionalism, clarity and detail in the prompts you provide. Beavers added that being polite to your AI chatbot "not only ensures you get the same graciousness in return, but it also improves the AI's responsiveness and performance."
Survey shows people are polite to AI out of fear
According to research conducted in December 2024 by Future, the publisher that owns TechRadar, about 67% of people who use AI are polite to it in the US, compared to 71% in the UK.
The survey of more than 1,000 people show that about two-thirds of people say they are impolite to AI due to brevity. Additionally, about 12% of respondents say they are polite out of fear of future consequences.
Gabe Hauari is a national trending news reporter at USA TODAY. You can follow him on X @GabeHauari or email him at Gdhauari@gannett.com.
Try Our AI Features
Explore what Daily8 AI can do for you:
Comments
No comments yet...
Related Articles
Bloomberg
30 minutes ago
- Bloomberg
Tencent-Backed MiniMax Releasing AI Tools to Take On Manus
Chinese upstart MiniMax is slated to roll out a pair of AI products that autonomously perform tasks on behalf of users, upping its game in the realm of agentic AI against the likes of OpenAI and Manus. The products include a general AI agent similar to that of Manus that solves everyday tasks, and another autonomous tool for video generation, according to a person with direct knowledge of the matter. The Shanghai-based company plans to roll out the services to the public this week, said the person, who asked to not be identified discussing private information.
Yahoo
33 minutes ago
- Yahoo
Pagaya Closes New PAID Deal of $600 Million, Capping Record Quarter for ABS Issuance
Over $2.3 billion raised across 6 ABS deals in Q2, reinforcing depth and resilience of Pagaya's multi-asset funding platform Over $6 billion of ABS and non-ABS capital commitments raised this year to fund all asset classes NEW YORK, June 18, 2025--(BUSINESS WIRE)--Pagaya Technologies LTD. (NASDAQ: PGY) ("Pagaya" or "the Company"), a global technology company delivering AI-driven product solutions for the financial ecosystem, today announced the closing of a new $600 million AAA-rated asset-backed securitization (PAID 2025-4), funding personal loans originated through its proprietary network. This deal was oversubscribed and upsized, with 23 investors participating, including several repeat investors. This transaction follows the Company's recent $300 million AAA-rated RPM deal and its inaugural AAA-rated POSH securitization, which created over $1 billion in funding capacity for new point-of-sale loan originations – reinforcing the Company's consistent funding execution across asset classes. With this deal, Pagaya has now raised over $2.3 billion in ABS transactions in Q2 2025, the largest quarterly total in the Company's history. Pagaya's personal loan ABS program has become one of the most active and widely-recognized issuance platforms in the capital markets, known for providing access to diversified personal loan assets at scale and with consistent performance. The continued cadence of transactions underscores institutional confidence in Pagaya's sourcing and tech-enabled underwriting capabilities. "This transaction reflects both the reliability of our ABS engine and the quality of our personal loan assets," said Sahil Chandiramani, Pagaya's Head of Capital Markets. "In a market where consistent access to capital is a real differentiator, we're proud to deliver back-to-back-to-back executions across auto, POS, and personal loans—at scale and with repeat investor support." "This record-breaking quarter is a reflection of the strength and precision of our best-in-class capital markets engine—and the exceptional team behind it," said Evangelos Perros, CFO of Pagaya. "We're proud to be executing at scale, across asset classes, and with consistency that's rare in today's environment. There's much more to come as we continue building our platform to power the next chapter of consumer credit." Since 2018, Pagaya has raised nearly $30 billion across 73 ABS transactions to fund loan originations through its platform, spanning personal loans, auto loans, and point-of-sale financing. About Pagaya TechnologiesPagaya (NASDAQ: PGY) is a global technology company making life-changing financial products and services available to more people nationwide, as it reshapes the financial services ecosystem. By using machine learning, a vast data network and an AI-driven approach, Pagaya provides comprehensive consumer credit and residential real estate products for its partners, their customers, and investors. Its proprietary API and capital solutions integrate into its network of partners to deliver seamless user experiences and greater access to the mainstream economy. Pagaya has offices in New York and Tel Aviv. For more information, visit View source version on Contacts Investors & Analysts ir@ Media & Press press@
Fast Company
34 minutes ago
- Fast Company
The hidden step most companies miss in their AI strategy
When I talk to executives about rolling out AI in their organizations, all conversations have a similar sentiment. Leaders say, 'We're going all in on AI,' only to follow up with, 'But don't let it touch any of our data.' This contradiction leaves teams wavering between ambition and caution, unable to act. This hesitation is understandable. After all, 'bad' data isn't just inconvenient—it's dangerous. The risks around privacy, data security, and corporate intellectual property are real. But if we want AI to drive better decisions, faster actions, and more personalized experiences, we need a new process: a way to prepare, trust, and govern the data we feed our models. That's where the AI data clearinghouse comes into play, a concept that I've been testing as I speak with Alteryx customers around the world. Think of it like a pre-flight checklist for your enterprise data. No matter how advanced the aircraft, pilots simply don't take off without running through numerous safety checks. In the same way, before your data 'takes off' into your AI environment, it gets inspected, validated, and approved. This shift turns AI from a boardroom buzzword into something teams can trust. HIDDEN COSTS OF BAD DATA To understand why a clearinghouse approach matters, let's look at what happens when organizations skip this critical step. Gartner reports that organizations will abandon 60% of AI projects unsupported by AI-ready data. And that's not surprising; even the most sophisticated AI can't transform bad inputs into valuable insights. These data problems cascade throughout an organization. They start small—perhaps a pricing recommendation that doesn't align with market realities or a customer insight that misses crucial context, but the ripple effects spread quickly. Consider the California car dealership whose AI system negotiated a $1 car sale. That's not just an amusing anecdote—it represents real dollars lost and trust eroded. The hidden costs multiply from there: decision paralysis as executives question every AI recommendation, expensive systems that gather digital dust, and perhaps most damaging, a cultural resistance that whispers, 'Maybe we should stick to spreadsheets and gut instinct after all.' That's why we need to approach AI implementation with data quality, not algorithm sophistication, as our north star. THE AI DATA CLEARINGHOUSE APPROACH How do we bridge this divide between AI ambition and data reality? Enter the AI data clearinghouse, a strategic framework that ensures only approved, accurate, and context-rich data powers your AI. Think of it not as another layer in your already complex tech stack, but as an essential trust-building solution that works alongside your existing CRM, ERP, and data lakes. This approach combines validation checks that catch problems early, business context layers that give meaning to raw numbers, cross-departmental approval workflows, and traceable documentation of data sources and changes. What makes this work is putting human expertise at the center. I saw this firsthand with a retail customer where the marketing and customer success teams defined 'churn' differently: one used the last engagement date while the other used the previous purchase date. Their AI model mistakenly flagged thousands of active customers as lost, triggering unnecessary reactivation campaigns that confused customers. What if curiosity about how your organization defines its metrics became the foundation of your AI strategy? Data preparation becomes an opportunity to create shared understanding rather than a technical hurdle. BUILDING TRUSTED AI FOUNDATIONS The journey to bringing the AI data clearinghouse to life in your organization begins not with technology, but with people. Before writing a single line of code, focus on building cross-functional teams that bring diverse perspectives. The most successful AI initiatives always start with genuine leadership buy-in. Only after establishing this collaborative foundation should you identify which business processes would benefit most from AI enhancement. Where could better decisions create the most value? Where do teams struggle with information overload or repetitive analysis? With clear objectives in place, focus on mapping the critical data sources that will fuel your efforts. This isn't just about identifying systems like Salesforce or Workday; it's about connecting with the people who truly understand what the data means. As you build your data workflows, weave in regulatory considerations from the start. GDPR and emerging AI-specific frameworks aren't checkboxes to tick off at the end. Rather, they are design principles that should shape your approach from day one. Perhaps most importantly, create feedback loops that allow your system to evolve. Effective AI implementations aren't static; they continue to improve as business needs change and teams discover new insights about their data. THE TRUST TRANSFORMATION Implementing AI is ultimately a trust exercise, and everyone in the company has slightly different motives. Executives want a competitive edge. Analysts want clarity. Compliance teams want control. And end users just want to be sure the answer they get is the right one. The companies that will succeed with AI aren't necessarily those with the most advanced algorithms or the biggest tech investments. Success will come to organizations that build thoughtful connections between their existing data environments and new AI capabilities.
