- Everyday AI
- Posts
- How the Future is Being Shaped by AI-Powered Autonomy
How the Future is Being Shaped by AI-Powered Autonomy
ChatGPT considers including ads, NSF and NVIDIA fund scientific AI push, consumer groups call out Grokās NSFW image tool and more!
Everyday AI
August 14, 2025
š Subscribe Here | š£ Hire Us To Speak | š¤ Partner with Us | š¤ Grow with GenAI
Outsmart The Future
Today in Everyday AI
6 minute read
š Daily Podcast Episode: How is AI-powered autonomy shaping the future of work? Discover the four big trends from Accentureās Technology Vision 2025 report and see why autonomy is redefining human expertise. Give it a listen.
šµļøāāļø Fresh Finds: Google Photos ads AI Create tab, Cohere hires Metaās former research lead and HTC to release AI glasses. Read on for Fresh Finds.
š Byte Sized Daily AI News: ChatGPT considers including ads, NSF and NVIDIA fund scientific AI push and consumer groups call out Grokās NSFW image tool. For that and more, read on for Byte Sized News.
š§ Learn & Leveraging AI: Weāre taking what we learned from Accenture, an industry leader, and breaking down how AI-powered autonomy is shaping your workplace. Keep reading for that!
ā©ļø Donāt miss out: Did you miss our last newsletter? We talked about the U.S. adding trackers to AI chip exports, GPT-5ās one model plan failing, Apple shifting toward AI hardware and more. Check it here!
How the Future is Being Shaped by AI-Powered Autonomy š
Work is changing from human-led to AI-powered autonomy.
How should we all prepare?
And how can we even trust an AI-powered workplace when most people can't even explain the basics of AI?
We're learning from the experts.
Accenture's Mary Hamilton joins the Everyday AI show to talk about building trust in an autonomous workplace, how we can prepare for the future of work, and four emerging AI trends you can't miss.
Don't miss this.
Also on the pod today:
ā¢ Autonomy and Enterprise AI Adoption š¢
ā¢ Agentic AI Models and Productivity Shifts š
ā¢ Continuous Learning Loops in Workplace AI š§āš«
Itāll be worth your 29 minutes:
Listen on our site:
Click to listen
Subscribe and listen on your favorite podcast platform
Listen on:
Hereās our favorite AI finds from across the web:
New AI Tool Spotlight ā Chance AI is AI-powered visual search, Tanka is an AI messenger with smart reply and Webdraw lets you explore and build AI apps with 50+ models.
Google ā Google Photos is adding a new Create tab to house all AI editing tools.
Google ā Google Flights can now suggest cheap flights with AI based on your description of where you want to go.
Cohere ā Cohere has hired Metaās former AI research lead as its new chief AI officer.
AI Tech ā HTC is coming out with AI glasses.
Money in AI - Balfour Beatty, a London based builder, is investing nearly $10M in Microsoft AI.
1. OpenAI Wonāt Rule Out Ads in ChatGPT ā But Is Cautious ā ļø
OpenAIās ChatGPT team is keeping ads on the table but stressing caution and tastefulness, with head Nick Turley telling Decoder the company wonāt ārule it out categoricallyā while emphasizing the strength of subscriptions and the need to preserve user trust.
According to Turley and Bloomberg figures, OpenAI projects $12.7 billion in revenue this year from subscriptions even as it burns cash and doesnāt expect positive cash flow until 2029, and ChatGPT recently topped 700 million users with about 20 million paid subscribers. The company is also testing āCommerce in ChatGPTā to earn referral cuts from purchases while insisting that affiliate deals wonāt skew recommendations.
2. NSF and NVIDIA Bankroll Open Scientific AI Push š¬ļø
According to the National Science Foundation, NSF is providing $75M and NVIDIA $77M to fund the Open Multimodal AI Infrastructure to Accelerate Science (OMAI), led by the Allen Institute for AI, to create open-source multimodal models tuned for scientific research.
The move is timely: as model development costs outstrip university budgets, this publicāprivate investment aims to give academics affordable, high-quality tools to speed materials, biology, and energy breakthroughs while training a broader AI-ready workforce.
3. Consumer Groups Demand Probe of Grokās NSFW āImagineā Tool š£
A coalition led by the Consumer Federation of America this week asked the FTC and all 50 state attorneys general to urgently investigate xAIās Grok āImagineā feature after The Verge found it could produce topless deepfake videos of Taylor Swift using its āSpicyā mode. The tool currently blocks āSpicyā for real photo uploads but still generates nude videos from AI-created images, a gap the groups warn could enable nonconsensual deepfakes and harm minors if loosened.
The letter ā signed by 14 consumer protection groups including EPIC and the Tech Oversight Project ā highlights xAIās pattern of removing moderation under freeāspeech rhetoric, raising immediate regulatory and safety concerns.
4. Metaās AI Chatbot Policies Spark Controversy Over Child Interactions š¬
A leaked internal document reveals that Metaās AI chatbots were once allowed to engage children with romantic and sensual language, sparking serious concerns about ethical boundaries, according to Reuters. While Meta confirmed the documentās authenticity and insisted these examples were errors, the incident highlights the challenges tech companies face in setting safe AI interaction rules.
The policies also reportedly permit generating false statements if labeled as such and allow some demeaning content based on protected characteristics, raising further questions about content moderation.
5. Google Pours $9B into Oklahoma Cloud and AI Infrastructure š°
According to Google, the company will invest an additional $9 billion in Oklahoma over the next two years to build a new Stillwater data center campus, expand its Pryor facility, and fund workforce and education programs tied to cloud and AI.
As part of a broader $1 billion U.S. education push, the University of Oklahoma and Oklahoma State University joined the Google AI for Education Accelerator for no-cost Google Career Certificates and AI training, while funding with the electrical training ALLIANCE aims to grow the electrical workforce by 135%.
6. AI-Designed Antibiotics Show Early Promise Against Superbugs š
A new study from MIT says generative AI has designed two novel compounds that killed drug-resistant gonorrhoea and MRSA in lab and mouse tests, marking a timely leap in antibiotic discovery as resistance drives over a million deaths annually.
The models trained on 36 million molecules built drugs atom-by-atom, avoided known antibiotic scaffolds, and filtered for human safety, though only two of 80 top gonorrhoea candidates were actually synthesizable. Researchers caution the compounds still need one to two years of refinement and the long, costly clinical-testing pipeline before any prescription use.
š¦¾How You Can Leverage:
High-end photo editing used to require years of training.
Adobe Photoshop demanded hundreds of hours to master one task like a seamless background removal.
Now?
You literally just say "swap out this background" and it automagically happens.
Same professional results. Zero technical expertise required.
Mary Hamilton said the same level of AI autonomy is happening everywhere.
And we've all gotta be prepared.
Mary is a Managing Director at Accenture, Leading the Connected Innovation Centers Globally. So, she knows a bit about workplace innovation.
Like photo retouching, she said many professions requiring decades of expertise just essentially became accessible to everyone because of large language models.
Mary just laid out the blueprint from Accenture's 25th annual Technology Vision report, laying out the path forward for AI-enabled workplaces.
Read the whole thing for free here, or a quick preview:
Development costs hit zero. Robots gained conversational intelligence. Trust-building determines who wins the autonomy race.
And most importantly ā the companies implementing these changes right now will dominate their industries within 18 months.
Here's what ya need to know:
1 ā Development Costs Hit Zero, App Building Goes Wild š
Enterprise software creation just became stupidly cheap.
Mary calls this the Binary Big Bang. Development costs that used to run six figures now cost basically nothing.
Here's what's happening.
Marketing teams are cranking out custom analytics dashboards during their lunch breaks. Sales folks build automated proposal systems between client calls. Operations peeps design workflow tools while waiting for Zoom meetings to start.
All functional applications. All built in hours instead of quarters.
Mary told us how Accenture's Fortune 500 clients are pumping out dozens of micro-applications daily. Traditional IT approval processes can't even keep up.
The game changer? Natural language replaced coding as the development interface.
When language becomes programming, every single employee becomes a potential builder. Companies that unlock this capability will absolutely steamroll competitors still waiting for IT tickets to get approved.
Try This:
Identify your team's five most annoying daily tasks right now.
Test Zapier Central or Microsoft Copilot Studio against each problem this week.
Document what gets automated versus what doesn't work. You'll quickly map your automation readiness and spot the natural builders on your team.
2 ā Robots Just Got Conversational Smarts š§
Physical automation became genuinely intelligent overnight.
Mary explained how large language models are giving robots human-like reasoning abilities. Instead of painful programming for specific tasks, workers now just tell robots what they want in plain English.
This changes everything about manufacturing.
Robots used to require complex coding for every single function. Now they understand context, adapt to changing situations, and work naturally alongside humans.
Mary shared how Accenture partnered with Keon Group to deploy AI-driven robots in warehouse operations.
The results? Faster order fulfillment, way lower costs, and massively improved worker safety.
But here's the kicker most companies are missing. This isn't about replacing humans with robots. It's about supercharging what humans can accomplish.
Robots handle the physical precision and endurance stuff. Humans bring creativity and complex reasoning. Together they absolutely crush either working alone.
Mary emphasized how multimodal AI gives robots three-dimensional understanding of real environments. They can see, process visual info, and make smart decisions instead of just following pre-programmed responses.
Try This:
Walk through your most labor-intensive processes and spot tasks requiring both physical work and real-time decisions.
Research robotics companies offering natural language interfaces for those specific applications.
Book demos this month. Early adoption creates serious competitive advantages before this becomes standard.
3 ā Trust Tech Beats Training Programs Every Time šŖ
Trust determines autonomy success way more than fancy technology.
Mary shared the real secret behind enterprise AI adoption. Trust gets built through tiny moments of successful interaction, not boring training sessions explaining how systems work.
Most companies are doing this completely backwards.
They're teaching employees about tokenization and large language model architecture instead of just letting them experiment safely. Mary actually coaches her AI assistants on communication style and personality, not just accuracy. She tells them to cut the snark and fine-tune responses.
This builds genuine working partnerships.
The winning approach? Verify first, trust gradually. Start with low-stakes scenarios where AI screwing up creates minimal damage but success shows clear value.
Mary described how Accenture built verification technology right alongside AI deployment. Knowledge graphs provide business context for data interpretation. Cross-checking mechanisms catch potential errors before they mess up operations.
These safeguards speed up trust development like crazy.
The competitive advantage compounds every month. Organizations building systematic trust get enterprise-wide adoption while competitors battle resistance and fear-based paralysis.
Try This:
Start trust journals with three employees using AI daily.
Have them rate response quality and track confidence levels for different tasks.
After two weeks, analyze which interactions build trust fastest. Design your rollout around natural trust patterns instead of forcing adoption down people's throats.
Reply