Welcome to the Global Advisors A.I Voices podcast. I'm Maya.

Today is Friday, the sixth of November 2026.

This is your morning catch-up on A.I. Please like and follow the show for your daily A.I update.

The recent midterm elections have introduced a new layer of political risk for the expansion of A.I infrastructure in the United States. TechCrunch reports that public opposition to data centers is increasingly driven by fears that these buildouts will enrich wealthy individuals while simultaneously increasing living costs for local communities.

This sentiment has translated into tangible political outcomes, with data center projects proving to be a vote loser across both major parties.

Meta executives have expressed concern that this political backlash could jeopardize both existing and future investments. Dina Powell McCormick, a Meta executive, noted that the underlying issue is broader public sentiment regarding A.I, with communities asserting their right to reject local infrastructure projects.

Meta currently operates data centers in 23 states and has five planned projects of one gigawatt or more, including a significant facility in Texas.

The situation in Texas is particularly acute. NLW, reporting for The A.I Daily Brief, states that Democrat Gina Hinojosa defeated incumbent Greg Abbott in the gubernatorial race, a result viewed as impactful for the A.I industry. Hinojosa tied data centers to broader economic issues, calling Abbott's tax subsidies a 'corruption tax' and stating she opposes how they take advantage of Texans.

NLW further notes that the incoming administration is expected to use policy to slow down data center buildout, including a moratorium until regulations are written and ending tax subsidies.

NLW explains that Texas has been a primary location for A.I infrastructure buildout due to cheap gas in West Texas, making the potential policy shift there a major concern for the industry. Despite these hurdles, Meta sees a path to monetization for its compute investments through leasing, even if consumer A.I adoption targets are not fully met.

This suggests that while the political landscape is becoming more hostile, companies are still finding ways to justify their massive capital expenditures.

In response to the same political headwinds, Elon Musk has proposed a radical alternative: moving data centers to space. After spending nearly half a billion dollars backing Republican candidates who lost in the midterms, Musk posted on X that data centers in space are looking pretty appealing right now.

He argues that space offers consistent sunlight and lower build costs, though this view ignores significant engineering challenges related to heat dissipation.

Musk's suggestion is not entirely new. In April, he predicted that the cost of deploying A.I in space would drop below terrestrial costs within a few years. While this remains a forecast rather than a measured result, it highlights how political and logistical hurdles on Earth are driving some of the industry's most prominent figures to consider extreme solutions for A.I infrastructure.

While infrastructure debates rage, the timeline for achieving Artificial General Intelligence, or AGI, is being compressed by some experts. Nathan Benaich, author of the State of A.I Report, predicts that AGI could be reached within the next 12 months. Speaking on the MAD Podcast, Benaich argues that models may soon outperform humans in specific professional domains like legal and finance when given a compressed amount of time.

Benaich defines the ability of models to outperform humans in any arbitrary domain, limited only by compute and synthetic data, as approximately AGI. He suggests that achieving this level of capability would accelerate the arrival of true AGI. This forecast contrasts with more cautious industry views, but it underscores the rapid pace of development in high-stakes professional fields.

As the A.I market matures, the hierarchy of business success is shifting. Nate Herk, an A.I automation expert, argues that the most valuable businesses are no longer those selling built products, like websites, but those selling human judgment, such as C-suite A.I strategy. This transition reflects a broader market trend where technical execution is becoming commoditized, while strategic oversight commands a premium.

Data supports this shift in value. An IBM survey cited by Herk shows that Chief A.I Officer adoption in large organizations rose from 26 percent to 76 percent, indicating a surge in demand for strategic leadership. Furthermore, PwC found that workers with A.I skills earn a 62 percent wage premium, underscoring the financial reward for those who can effectively integrate A.I into business processes rather than just building the tools themselves.

Herk's ranking methodology evaluates businesses based on technical requirements, market space, and long-term durability. He forecasts that the window for building these high-value businesses is narrowing as market saturation increases, urging entrepreneurs to focus on areas where human judgment remains irreplaceable.

Google DeepMind has released Gemini Omni 1.2 Flash, a new model focused on fast and cost-effective video generation. The model is priced at 0.07 dollars per second for 720p video, representing a 30 percent price reduction compared to its predecessor, Gemini Omni 1.1 Flash. This update is designed to make high-quality video generation more accessible for developers and businesses.

Gemini Omni 1.2 Flash supports the generation and editing of videos up to 15 seconds long, increasing the limit from the previous capacity. The release introduces native video creation and multi-turn conversational editing capabilities. Additionally, the model provides stronger adherence to image and video references to maintain consistency for products, characters, and brand details, addressing a common challenge in A.I-generated media.

OpenAI has released GPT-Image-2.6 across ChatGPT, Codex, and the API, featuring two new models: Flare and Sunburst. Flare is optimized for speed, capable of returning images in as little as 6 seconds at low quality, while Sunburst targets higher quality and consistency. This dual-model approach allows users to choose between rapid generation and detailed output based on their needs.

The update includes several new features, such as GIF output, SVG export, markup-guided edits, and a new complexity parameter. These additions expand the utility of the model for web development and design workflows. Image output tokens for GPT-Image-2.6 Flare are priced at 30.00 dollars per 1M tokens, reflecting the cost structure for high-volume API usage.

Law firm A and O Shearman is committing 1.1 billion dollars over five years to accelerate its artificial intelligence strategy. The majority of the investment involves each practice committing at least 3 percent of its time to developing specialist A.I tools and services within their specific area of law.

This approach emphasizes embedding specific practice expertise rather than using generic tools, building on the firm's 2022 rollout of generative A.I, which was the first by any law firm globally.

Strong governance underpins the A.I use, with qualified lawyers retaining responsibility for advice and A.I governance committees overseeing deployment. This ensures that while A.I tools are developed internally, the final legal advice remains under human control. The firm's strategy reflects a broader trend in professional services, where A.I is being integrated into core workflows to enhance efficiency and expertise.

Continuing the discussion on A.I business durability, Nate Herk identifies personal branding as the most resilient model in the face of rapid technological change. Herk argues that while tools and platforms evolve, an established audience provides a direct investment that allows entrepreneurs to pivot more easily.

He scores personal branding and A.I education as a five for durability, noting that spin-off businesses can be powered by the trust already built with the audience.

Herk discloses that A.I education and personal branding is his main business, acknowledging potential bias in his ranking. However, he emphasizes that this model's strength lies in its independence from specific software vendors. By focusing on education and community, entrepreneurs can maintain relevance even as the underlying A.I tools shift.

Crucially, Herk warns that proof of work and trust must precede brand building. He advises against early distraction with branding efforts, suggesting that entrepreneurs should first demonstrate competence and reliability. Only then can a personal brand serve as a durable asset that withstands the volatility of the A.I market.

Hardware startup Aina has released a smart ring designed to address the 'bottleneck' of using a phone to approve A.I agent actions. The Aina Ring features a small display, a joystick-like button for quick acceptance or rejection, a mic, and a speaker, allowing users to talk to A.I agents, take notes, and make phone calls.

The device is available for preorder at 399 dollars for a limited early-bird period, with the price set to rise to 499 dollars when shipping begins in January 2027.

Aina raised 5.5 million dollars in July and positions the ring as a controller for both A.I and non-A.I applications. Apoorv Shankar, associated with Aina, notes that unlike previous failed gadgets like Humane and Rabbit, Aina does not aim to replace the phone but rather to act as a means to stop users from being a bottleneck so A.I agents can work on their behalf.

The Aina Ring competes with other smart rings like Pebble Index 01, Sandbar Stream, Oasis, and Natura, differentiating itself with added controls like a joystick, screen, and speaker.

A new United Kingdom-backed startup, Potential Sciences, is launching with the aim of accelerating drug discovery by bypassing traditional lab experimentation. The company plans to build chemical datasets by simulating atomic and molecular interactions using physics principles.

WIRED reports that the goal is to train A.I models on this simulated data to identify targeted drugs, potentially speeding up the plodding process of developing new medicines.

Potential Sciences launched on November 5 with seed funding in the tens of millions from SovAI, the United Kingdom government's A.I venture capital fund, and Twin Path Ventures. The precise figure or valuation was not disclosed. The company has rebuilt quantum chemistry software to run on GPUs, claiming this makes calculations up to 100 times cheaper than current methods.

To support these computationally intensive simulations, the company will access computing power through the United Kingdom's A.I Research Resource, including the Isambard-A.I supercomputer in Bristol. However, the startup has not yet settled on a revenue generation method or decided whether to make its models freely available, citing potential misuse concerns.

That's A.I Voices for today. I'm Maya. Thanks for listening.

Follow the show on Spotify, YouTube, Apple Podcasts, Amazon Music, or your favourite podcast platform, and explore the original voices and sources in the episode notes. See you tomorrow.