Welcome to AI Decoded, Quick Firm’s weekly publication that breaks down crucial information on the earth of AI. I’m Mark Sullivan, a senior author at Quick Firm, overlaying rising tech, AI, and tech coverage.
This week, I’m specializing in Meta’s bid to deploy superior industrial AI brokers throughout its large messaging community. I additionally have a look at Microsoft CEO Satya Nadella’s artful protection of information heart water utilization, and at a dramatic shift in public sentiment towards new information heart initiatives.
Signal as much as obtain this article each week by way of e-mail here. And in case you have feedback on this problem and/or concepts for future ones, drop me a line at [email protected], and comply with me on X @thesullivan.
A have a look at Meta’s new AI brokers for companies
Meta mentioned Wednesday that it’s going deeper into the enterprise of promoting synthetic intelligence brokers to firms of all sizes. The corporate announced new brokers for companies that can run on WhatsApp, Instagram, and Messenger. The brokers can deal with buyer assist requests, e-book calendar appointments, reply often requested questions, qualify leads, and escalate difficult queries to human operators. Over time, they may virtually actually evolve to finish extra advanced duties on behalf of companies.
Meta can be launching a Enterprise Agent Platform that can let companies construct and handle customized AI brokers. The platform connects to lots of of non-Meta programs, together with Shopify, Zendesk, and Shopee, the place these brokers might be deployed. The extra duty and autonomy a enterprise provides an agent, the higher the prospect that agent will get caught or makes errors on the client’s behalf. Safety publicity will increase, too. Meta says the platform affords bigger companies enterprise-grade controls, guardrails, and measurement instruments.
It’s straightforward to be cynical about Meta’s renewed forays into state-of-the-art AI fashions. Mark Zuckerberg has by no means met a brand new tech paradigm he didn’t need to lead. (Howdy, metaverse.) Beginning final summer season, Meta rebuilt its AI analysis group round a brand new (and really costly group) of researchers known as Meta Superintelligence Lab. The corporate could have the expertise to compete with Google, OpenAI, and Anthropic, however the query has been, “To what finish?” Does Meta want big, cutting-edge AI fashions to energy its core promoting enterprise? That’s unclear. But it surely in all probability does want state-of-the-art fashions to energy more and more autonomous brokers for companies.
For Meta, promoting AI brokers to companies may make industrial sense. The corporate can use the huge attain of its social networks (the place hundreds of thousands of companies have already got a presence) to seek out new prospects for its brokers. For some companies, particularly small ones, it might make sense to run AI brokers on the identical platforms the place they run adverts.
Many social community adverts already comprise a “click on to contact” button that opens a messaging session between a shopper and a enterprise, or maybe a enterprise’s agent. Meta has additionally been pushing companies to deal with messaging platforms like WhatsApp as assist channels, as prospects more and more choose texting to cellphone calls.
No matter you concentrate on the relative worth of Meta’s social platforms, they’ve been a boon for a lot of small companies. Meta’s brokers—in the event that they work properly and don’t alienate prospects—could also be a wonderfully rational extension of the enterprise aspect of social media.
Did Satya Nadella put the information heart ‘water debate’ to relaxation? Not precisely
Microsoft’s CEO, Satya Nadella, tried to defuse the rising controversy over AI information heart water consumption at Microsoft’s Construct convention this week, claiming that fashionable information facilities depend on extremely environment friendly, closed-loop cooling programs that require minimal replenishment.
Nadella’s remarks gained traction amongst AI accelerationists on X after they were shared by the AI commentator Alex Volkov, who urged Nadella had successfully put the “water debate” to relaxation.
Nadella mentioned that when these closed-loop programs are initially crammed, their annual water footprint drops considerably. To place the size in context, Nadella mentioned a contemporary information heart’s whole water utilization over a complete yr is roughly equal to that of a single restaurant. (The enterprise capitalist and Trump ally Marc Andreessen quipped: “And that is why we should additionally abolish eating places.”)
Not so quick.
Business analysts say Nadella is true relating to really closed-loop information heart cooling programs. Hyperscalers like Microsoft, Google, and AWS are aggressively shifting to superior closed-loop and direct-to-chip liquid cooling for brand spanking new frontier-model clusters. Microsoft said in 2024 that it could start constructing solely closed-loop programs that recycle water and use direct-to-chip cooling.
However a considerable variety of the world’s operational information facilities nonetheless depend on older evaporative cooling strategies that may eat hundreds of thousands of gallons of water a day. Many hyperscale information facilities constructed over the past decade use evaporative cooling towers, which repeatedly eat water (by way of evaporation) to dump warmth. Google, for instance, recently said that about two-thirds of its information facilities nonetheless depend on evaporative cooling. This water just isn’t reused indefinitely and have to be replenished.
So Silicon Valley can’t really absolve itself of water abuse till the hundreds of information facilities now supporting AI have transformed to water-efficient cooling.
Survey: Public opinion has swung 49 factors towards information facilities in simply 9 months
Public opposition to information facilities is surging because the bodily footprints of the AI increase mark the panorama, and native residents and politicians weigh the technological advantages towards the financial and ecological prices. In response to a new poll carried out by Embold Analysis, 71% of People now oppose having a knowledge heart constructed close to their houses. This represents a dramatic 49-point swing in web opposition since September 2025, when the general public was roughly evenly break up.
The backlash crosses political and demographic traces, with opposition reaching 80% amongst younger voters and 63% amongst Trump voters. The info displays intensifying native resistance over energy grid pressure, water utilization, and noise air pollution. The friction is already impacting infrastructure rollouts, with public opposition scuttling at the least 20 initiatives within the first quarter of 2026 alone.
Extra AI protection from Quick Firm:
- Trump’s AI order gives Washington a look at frontier models, but not much leverage
- Coursera now offers an AI-powered feed of short form educational content
- OpenAI CEO Sam Altman makes a lot of predictions. Here’s how they’ve fared so far
- Your AI-dar probably doesn’t work
Need unique reporting and development evaluation on expertise, enterprise innovation, future of labor, and design? Sign up for Quick Firm Premium.
