May brings AI into more of the places where business work happens: across languages, inside connected software, and on a phone between appointments. Greater reach is useful only if it improves a task without creating extra correction or confusion. The central business question is therefore not how many channels can access AI. It is whether the organization can preserve accurate information, consistent procedures, and a clear approval step as access becomes easier.
1. PaLM 2 highlights multilingual business possibilities
Google introduced PaLM 2 on May 10, describing improved multilingual, reasoning, and coding capabilities. The company presented a family of model sizes and announced uses across its products. These are provider-reported capabilities and product plans; performance on a company's own terminology still needs to be established.
For a business serving customers in more than one language, the potential is broader than literal translation. An assistant might help prepare alternate explanations, compare phrasing, or simplify an internal instruction. Yet product names, contractual terms, and cultural context can change the meaning of a seemingly fluent translation. Important customer communications need review by someone qualified to judge the language and subject.
Business takeaway: Build a small approved glossary before testing multilingual assistance. Use it to evaluate a handful of common messages and record where the tool fails to preserve meaning. Measure the review time as well as the drafting time. The goal is a repeatable communication process, not simply a greater volume of translated text.
Source: https://blog.google/innovation-and-ai/products/google-palm-2-ai-large-language-model/
2. Microsoft focuses developers on copilots and connections
At Build on May 23, Microsoft outlined developer tools and an expanding plugin approach for AI assistants. The announcement described plugins as connections that can let AI systems work with APIs, current information, and business data. It also announced plans to adopt the plugin standard introduced for ChatGPT. The package included announcements and previews rather than one universally available finished product.
The business implication is that useful assistance increasingly depends on integration. A model that can explain a procedure is different from a system that can retrieve the correct record or initiate a business action. Once actions are possible, permissions, validation, and confirmation become part of the application design.
Business takeaway: Separate reading from changing. For a first connected pilot, allow the assistant to retrieve approved information and prepare a proposed action. Have a person approve any change to a customer record, purchase, or outgoing message. Ask the implementer to demonstrate how an incorrect suggestion is rejected before it can affect another system.
3. ChatGPT arrives on iOS with voice input
OpenAI released a ChatGPT iOS app on May 18, beginning its rollout in the United States. The app synchronized conversation history and used Whisper for voice input. Plus subscribers could access their associated GPT-4 capabilities. The announcement described expansion to other countries and an Android app as upcoming work.
For an owner or field employee, mobile access could help capture a generic idea or prepare an outline while away from a desk. Voice input also creates a practical risk: spoken notes can include names, private customer details, or background information that should not be sent to an unapproved service. Convenience does not decide whether the information belongs there.
Business takeaway: Define a mobile use case with a clean handoff. For example, staff can turn a non-sensitive observation into a draft checklist, then review and store the approved checklist in the company's normal system. Avoid making a personal chat history the only record of an important instruction or customer commitment.
Source: https://openai.com/index/introducing-the-chatgpt-app-for-ios/
The management decision
The common challenge is maintaining a single business process across multiple interfaces. A phone, a browser, and a connected assistant should not each become a separate version of company policy. Select the authoritative place for approved information and make the human handoff explicit. A pilot succeeds when employees can complete a useful task with fewer steps and no loss of accountability. Track abandoned drafts and repeated corrections, too; they show where apparent convenience may be adding work rather than removing it.
Present-day invitation
Want to turn these ideas into a practical improvement for your business today? Book an appointment with Brice Prather at https://loyti.com/book to discuss AI, custom software, or business automation. Bring a question, a costly bottleneck, or a growth opportunity. Together, we can identify a sensible next step, what it would require, and how its value should be measured.