Sunday, August 9: no same-day launch verified; two weekend drops

The Sunday check found no qualifying AI tool or product release dated August 9 that could be verified before this briefing. Two nearby releases are still useful because their boundaries are clear: OpenAI bringing a presentation startup's team into ChatGPT, and Rippling turning AI usage into a finance and governance problem.

OpenAI acquired NextSlide, whose team is now helping build ChatGPT. NextSlide's founder, Ahmed Beshry, describes the startup's product as turning prompts, notes, documents, or research into polished, editable presentations. TechCrunch reported the announcement on August 8 and notes that the financial terms were not disclosed. 12

This is an acquisition rather than a new standalone tool release. The useful question is whether that presentation workflow becomes part of ChatGPT; the announcement does not promise a specific feature or release date.

Rippling's official launch post dates AI Spend Console to August 6. It gives finance and technology leaders a way to see AI costs, connect them to business outcomes, and govern approved large language model use. The company says the product is available with a 30-day free trial and does not require an existing Rippling subscription. 3

The product connects Anthropic, OpenAI, and Cursor; breaks spending down by employee, team, or department; and maps token costs to signals such as pull requests and code reviews. It also supports plain-language questions and dashboard creation without SQL. Rippling's AI Gateway, for model routing and spend limits, is still listed as coming soon. 4

TechCrunch reported that Rippling's own AI token spending had reached 40% of its research-and-development headcount budget, including one engineer spending about $50,000 a month. The company said caps and cheaper model routing reduced that share to about 15% while usage stayed high; these are company figures, not an independent audit. 5

Together, the releases point to two different jobs: generating the finished presentation and measuring whether an organization's AI spending is producing enough value to keep. Neither announcement removes the need to test the workflow against real work.