BIG TECH

How Amazon launched a data center charm offensive

Data centers don't have a great reputation among the public. Amazon is making a new move to fix that. 

On Friday, Amazon announced "Built Together," an investment framework for building and running data centers in communities in response to the growing public backlash against these facilities. The initiative includes a $1 billion investment over the next five years to fund education, job training, energy affordability, and water preservation in the communities where its data centers exist. 

Amazon's plan to appease these communities includes three central pillars: 

• Creating education and workforce opportunities by investing in free community college and "hands-on" training for residents in its data center communities. The company is specifically funding associate degrees in "high-demand" fields, including electricians, HVAC experts, fiber opticians, IT, healthcare, education, public safety, and advanced manufacturing. 

• Ensuring affordable energy and clean water solutions by creating grants for energy efficiency upgrades to K–12 schools, community buildings and homes in these communities, covering the costs to install heat pumps, HVAC systems, insulation, water heaters, batteries and solar installations, reducing energy costs by up to 40%.

• Funding local-specific, high-priority investments with "communities in the lead," such as affordable housing, food security, disaster preparedness, or whatever specific needs each individual community has. 

In the announcement, Matt Garman, CEO of AWS, cast doubt on a number of common narratives around data centers, including that they use excessive water, drive up energy rates, emit large amounts of pollution and bring nothing to communities. He rejected the data center moratoriums that are up for consideration in a number of states, put forth by politicians from both sides of the aisle, which aim to clamp down on the AI infrastructure buildout due to strain on the energy grid and environmental impacts. 

Garman said that, if enacted, "The US could be writing its own losing ticket to this [AI] race, and the consequences would last generations. As a country, we can’t afford to find ourselves in that position." 

However, in defending data centers, Amazon will have to fix a behemoth of a PR problem. These facilities are largely reviled, with one poll from Gallup showing that an average of 7 in 10 Americans oppose them being built in their communities. Additionally, these facilities have also become a scapegoat for broader anxiety and fear around AI, creating a rallying point that extends across political parties or ideological differences.

Amazon has a lot riding on rehabilitating the image that data centers have in the public eye. The company has long been a kingpin in the cloud space, with AWS representing roughly a third of the global cloud services market share. The AI buildout has only raised the stakes, exacerbating the need for a rapid buildout into communities that are wary at best, and deeply, loudly opposed in many cases. Though this initiative aims to help appease these mistrustful communities, Amazon may have another factor to contend with: The ever-growing AI doomerist narrative. With many of AI's leaders claiming that the tech could trigger catastrophic risks that decimate national security and put lives at risk, those ideas have permeated the zeitgeist and struck fear into a public that's already underinformed about the technology. That alarmism could undermine Amazon's mission, no matter how many free HVACs it funds.

TOGETHER WITH NOOKS

Sales is harder than code. Nooks Labs is working on it.

Coding agents get feedback in seconds: the tests pass or they fail. Nooks learned that sales agents can wait weeks to find out whether a move worked. Much of what matters, like who's really making the decision, never makes it into the CRM.

Nooks' agents scored 97% on the company's own eval set, but reps still double-checked every output. So Nooks Labs, the company's applied AI team, built a harder benchmark from real deal data. The team then fixed what it exposed with calibrated confidence scores, a way to test a strategy against past deals before it runs, and 500+ examples of bad AI writing used as guardrails.

Today, 96% of agent-written emails go out without edits. The open question now: what does it take to go from 2x productivity gains to 10x?

Discover the research →

PRODUCTS

New voice model shows Microsoft's AI strategy

Microsoft once relied nearly entirely on OpenAI's frontier models to power its cutting-edge tools. But now it's increasingly launching its own models.

On Thursday, Microsoft unveiled its first-ever streaming transcription model, MAI-Transcribe-2-Streaming, offering low-latency, real-time transcripts in 60 languages. At launch, it topped the Artificial Analysis Word-Error-Rate Leaderboard, which measures the percentage of words transcribed incorrectly in the final transcription, beating out models from Grok, Meta, and OpenAI. 

Microsoft claims that the model can deliver text as soon as 100 milliseconds after receiving audio. In AI applications, it allows models to begin reasoning sooner, enabling them to begin tasks before the user has finished their thought.

The model also helps with low-latency, real time subtitling. Microsoft claims that internal evaluations found it to be twice as fast as its closest competitors for use cases such as real-time dictation or subtitling. 

Additionally, Microsoft launched two new voice models: 

• MAI-Voice-2.1: Strongest multilingual text-to-speech model, with expanded support to 23 languages and 26 locales. 

• MAI-Voice-2.1-Flash: Has the same language support as the model above, but has also been optimized for high-volume, latency-sensitive workloads. 

As voice models start to power seamless interactions with voice agents across AI software and hardware, Microsoft isn't the only company with its eye on the space. Suno, best known for its music-generation models, has now expanded beyond its core offering with the release of Speech, which it calls the first audio model to generate voice and music in a single, cohesive track, the company announced on Thursday.

The way Suno Speech works is that users can create spoken audio set to background music by typing text, describing the voice and musical style, and then the model outputs a reading in a voice that fits the user's description with original music to go along with it. It is available in beta on the platform, with the company caveat that, since it is still in beta, it doesn't always perform accurately, for instance, noting that British accents drift into Australian, and dramatic pauses may sound overdone.

Microsoft's latest model releases are all domain-specific: MAI-Thinking-1, Microsoft AI's first reasoning model; MAI-Code-1.1-Flash, a model focused on coding assistance; and MAI-Image-2.6, its most advanced image model. Microsoft has several reasons to stick with this strategy. By focusing each model on a single domain, it can concentrate its resources on making that model high-quality and high-performing, as these smaller models' strong benchmark results show. Building general models that cover a broad range of use cases is very resource-intensive. The approach also suits Microsoft's business and enterprise customers, where the company has its best opportunity to win since enterprise users rely on AI for specific tasks in their everyday workflows. For example, for an engineer, a cutting-edge coding model matters more than a general model that can tell them weather patterns and do math.

TOGETHER WITH JUMPCLOUD

AI agents need admin access, too. Just not all the time.

Some AI agents will need privileged access to restart servers, connect to databases, or manage infrastructure. That doesn’t mean they should carry permanent admin credentials.

The same security principles used for human admins should extend to agents: access should be scoped, governed, auditable, and revocable. As agents take on more sensitive work, Agentic IAM gives IT a framework for governing who or what is acting and what it’s allowed to do.

Secure every identity. Human or not.

GOVERNANCE

Can guardian agents make enterprise AI safer?

AI agents can act faster than humans can review them. So how do companies keep control without slowing everything down?

In this episode of The Deep View Conversations, Jason Hiner sits down with Blake Brannon, chief innovation officer and a co-founder of OneTrust, to explore why AI governance needs to evolve from static rules to decisions made in real time.

Brannon explains why an agent can inherit an employee's identity and access without inheriting their trust. He makes the case for independent guardian agents that evaluate what other agents are trying to do, apply a company's policies, and redirect or block actions when necessary.

One example captures the challenge: deleting CRM records could be routine cleanup or an attempt to inflate a salesperson's performance. The action alone doesn't tell you whether it should be allowed. Its purpose matters.

Topics covered:

• How OneTrust evolved from data privacy to AI governance
• Why first-party data and customer trust go together
• Why human review can't keep pace with a growing population of agents
• How guardian agents and independent checks could help close the trust gap
• The difference between containing AI and aligning it with business policies
• Why governance should help companies move faster
• What the bring-your-own-device era teaches us about AI adoption
• How Blake uses visual storytelling and voice-to-text tools to work more effectively

If you're deploying AI agents and trying to move beyond experiments without giving up control, this conversation offers a practical way to think about the systems that need to surround them.

📺 Watch on YouTube
🎧 Listen in your favorite podcast player

Subscribe to Deep View Conversations for interviews with the leaders shaping the future of AI, business, and technology: tdv.transitor.fm

LINKS

• Apple tightens "Full Disk Access" controls amid AI agent risks

• ArXiv caps submissions amid record submissions fueled by AI

• Dutch retailer Hans Anders stops Meta Ray-Ban sales over privacy concerns

• Prosecutors arrest California man for smuggling Nvidia chips to China

• FieldAI, a robotics startup, raises $700 million at $10 billion valuation

• Nvidia, SoftBank each invest $10 billion dollars in final OpenAI round

• Meta will let hardware makers build their own versions of Muse gadgets

• Grok Bot: hand off coding tasks to Cursor, manage PRs with GitHub and Origin plugins

• Google Stitch: Now has a Stitch CLI for connecting to local coding agents, and more

• Project Continuum: Runway Labs, a new operating system research application 

• Notion: launched token sharing with ChatGPT

• Deloitte: Google AI Architect

• Clove Dental: AI Operations

• Noscen: Founding AI Engineer

• Booz Allen Hamilton: Agentic AI and Data Engineer

GAMES

POLL RESULTS

Are you concerned about AI's impact on education?

Yes (73%)
Somewhat (16%)
No (7%)
Other (4%)