Nội dung
- Kimi Work Goal wants to turn AI into a 24/7 “employee”
- Terafab: a giant chip infrastructure move for the AI era
- OpenAI continues to bet on Rust and the systems ecosystem
- Tria shows crypto has moved beyond the user-testing stage
- Noam Shazeer leaving Google to join OpenAI heats up the talent race
- Midjourney continues on its own path, without relying on investors
- Midjourney Medical opens up healthcare ambitions through full-body scanning technology
- Sam Altman jokes about “Noam,” but the real message is fierce talent competition
- OpenAI publicly welcomes Noam Shazeer as a strategic milestone
- Elon Musk’s remarks continue to pull AI into social debate
- When technology, law, and public trust collide
- A perspective for the Vietnamese market
- References
Last week, the global AI landscape continued to revolve around three major themes: software agent capabilities, the race for top talent, and the challenge of compute infrastructure. For Vietnamese marketers, this is not just a technology story, but also a signal of how tools, content, and digital operating models will change very quickly in the near future.
The roundup below brings together the 11 most notable updates, from new AI platforms and OpenAI’s moves to Midjourney, as well as information related to chips, the open-source ecosystem, and social topics that are stirring discussion across the tech world.
Kimi Work Goal wants to turn AI into a 24/7 “employee”
Kimi.ai says Goal Mode in Kimi Work is designed so the desktop agent can run continuously until a task is completed, making it suitable for long-running, multi-step workflows. This is a notable direction because AI is shifting from a response tool to an agent capable of pursuing goals.
For marketing teams, this suggests a new layer of applications: project tracking, data collection, report preparation, or repetitive workflows without humans having to “remind” the system at every step. However, the higher the level of automation, the more businesses need clear quality-control processes and data-access permissions. Source: Kimi.ai on X.
Terafab: a giant chip infrastructure move for the AI era
Information circulating online suggests Terafab is being described as the world’s largest chip factory, tied to the ambition of producing AI compute capacity at an extremely large scale each year. Source data says the project aims to reach 1 terawatt of AI compute per year, with an initial phase cost of around 25 billion USD and a total figure that could reach 119 billion USD.
While these figures still need to be tracked further through official verified sources, the market message is already clear: the AI race is not only about models, but also about the semiconductor supply chain, chip manufacturing, and the energy needed to power compute. For marketers, this is a reminder that every AI service they use depends on massive physical infrastructure behind the scenes. Source: a compiled post on X citing related remarks from Muskonomy and information from Trump/Musk.
OpenAI continues to bet on Rust and the systems ecosystem
Greg Brockman said OpenAI will contribute 600,000 USD to the Rust Foundation, including Platinum membership and additional support for maintainers in the community. The accompanying message is consistent: OpenAI sees Rust as the future of high-performance software systems.
For digital product teams, this move shows that the AI competition is not only about bigger models or better prompts, but also about the ability to build infrastructure that is safe, stable, and resource-efficient. Rust is increasingly being mentioned as an important foundation for the system layers behind AI. Source: Greg Brockman on X, quoting Charlie Marsh.
Tria shows crypto has moved beyond the user-testing stage
In the source data, Tria is described as having reached 100 million USD in transactions, 500,000 users, and more than 30 million USD in card spending in less than 6 months. The notable point is that the product emphasizes self-custody, runs on Visa, supports more than 150 countries, and handles multi-token, multi-chain payments.
Although this is more of a Web3 ecosystem story than a pure AI one, it is still relevant to the tech market because many AI applications are gradually integrating payments, digital identity, and cross-border experiences. For marketers, it is a clear example of how “real utility” remains the decisive factor in adoption. Source: OG FIGO on X.
Noam Shazeer leaving Google to join OpenAI heats up the talent race
Sam Altman confirmed that Noam Shazeer has joined OpenAI and called him one of the people he wanted to work with from very early on. Noam Shazeer also shared that leaving Google was a difficult decision, but that he is proud of his former team and what they built.
This development shows that the AI battle has shifted strongly to the front line of talent acquisition: people who deeply understand model architecture, systems, and products. For Vietnamese businesses, the lesson is that it is not enough to look only at tools; teams must be strong enough to use AI sustainably. Source: Sam Altman and Noam Shazeer on X.
Midjourney continues on its own path, without relying on investors
Source data describes Midjourney as a community-funded research lab with no outside investors, using revenue from its image-generation product to fund all R&D. The figures cited show the company grew very quickly in its first year, while also pursuing 8 active projects split evenly between hardware and software.
What stands out is that Midjourney is not stopping at image generation, but is also expanding into hardware products, a sign that leading AI companies are looking to move beyond the software layer. For marketers, this is an example of the power of product branding: real value can fund its own growth if the product is distinctive enough. Source: Nick St. Pierre on X, based on comments about David Holz and Midjourney.
Midjourney Medical opens up healthcare ambitions through full-body scanning technology
Midjourney is said to have announced a new branch called Midjourney Medical, highlighted by the idea of an ultrasound CT full-body scanner. The source description says the device is designed for safe scanning without X-rays, magnets, or radiation, and can be repeated many times.
If this move goes further, it would be a rare example of a company known for creative AI stepping into healthcare by combining hardware, imaging, and user experience. Even though it is still at the concept stage, it shows how quickly the boundaries of AI applications are expanding. Source: Nick St. Pierre and Robert Scoble on X.
Sam Altman jokes about “Noam,” but the real message is fierce talent competition
Sam Altman posted a humorous line about being “unable to explain why people named Noam are so good at AI,” while also expressing excitement about having another Noam Shazeer at OpenAI. Although playful, it is still a remark that shows how lively and competitive AI hiring culture has become.
Behind the light tone on social media is a race to attract people who understand models at a very deep level. For marketers, this reflects a reality: the bigger AI becomes, the more human talent turns into a strategic advantage. Source: Sam Altman on X, quoting a response about Noam Shazeer.
OpenAI publicly welcomes Noam Shazeer as a strategic milestone
In another post, Sam Altman said Noam Shazeer was someone he wanted to work with from the earliest days of OpenAI and that it “finally took 10 years.” Noam Shazeer, meanwhile, expressed excitement about the new team and thanked Google for his previous time there.
This chain of statements reinforces the message of a talent reshuffle at the very top of the AI industry. For businesses, this is a time to pay closer attention to retention strategy, culture building, and creating an environment attractive enough for top tech talent. Source: Sam Altman and Noam Shazeer on X.
Elon Musk’s remarks continue to pull AI into social debate
In a series of posts related to a report on serious incidents in the UK, Elon Musk continued to make attention-grabbing remarks such as “this is just the tip of the iceberg” and “no one can save you when they become the police.” Although not directly AI news, statements from major tech figures still tend to spread widely across digital spaces.
For communications professionals, this is a reminder that the tech ecosystem and social debate are becoming increasingly intertwined. A statement from a tech leader can quickly move into politics, public trust, and brand crisis territory. Source: Elon Musk on X and the reposted content.
When technology, law, and public trust collide
The final content sequence in the source continues to revolve around reactions to investigative reporting and comments on the role of public authorities. Although it is not a product announcement or model launch, it shows that the media environment around Big Tech today has gone far beyond a technical story.
For tech brands, the lesson here is that public messaging always has consequences that spread far beyond the original intended scope. Every statement, especially from highly influential figures, can affect public perceptions of AI, data, and social responsibility. Source: Elon Musk on X and the reposted content.
A perspective for the Vietnamese market
For Vietnamese businesses and marketers, these 11 updates point to one thing: AI is moving very quickly from a “support tool” to an “operating infrastructure.” That means marketing teams not only need to know how to use AI for writing, image generation, or data analysis, but also need to understand how to automate workflows, control output quality, and choose tools that fit business goals.
At the strategic level, Vietnam should pay special attention to three trends: long-running AI agents like Goal Mode; the chip and compute infrastructure race; and the shift in talent and the open-source ecosystem. Companies that build a structured way of working with AI early will have a clear advantage in speed, cost, and scalability over the next 12–24 months.
More importantly, this roundup shows that AI does not stand alone, but is intersecting with healthcare, payments, communications, and hardware. That is also why Vietnamese marketers need to approach AI as a broader competitive capability, not just as a content-generation tool.
See more marketing news and guides at https://marketing365.vn.
Follow more updates from Marketing365 to stay up to date with the latest marketing trends.
Read more articles in the same category AI developments.
This article focuses on AI news with a perspective for the Vietnamese market.
References
- Kimi.ai — Introducing Goal Mode in Kimi Work Goal
- Muskonomy — Terafab chip factory update
- Greg Brockman — Rust Foundation commitment
- OG FIGO — Tria adoption update
- signüll — Google/OpenAI Noam Shazeer satire thread
- Nick St. Pierre — Midjourney overview thread
- Robert Scoble — Midjourney Medical comment
- Sam Altman — Noams and AI
- Sam Altman — Welcome Noam Shazeer to OpenAI
- Elon Musk — statement on UK report
- Elon Musk — police-related statement
