- Everyday AI
- Posts
- Ep 723: From AI Chatbot to Autonomous Coworkers: How Consumer AI Has Changed and What's Next
Ep 723: From AI Chatbot to Autonomous Coworkers: How Consumer AI Has Changed and What's Next
OpenAI Announces largest-ever funding round, 300+ Google and OpenAI Employees Back Anthropic in Open Letter, Meta Signs Multi-Billion Dollar TPU Deal With Google, and more
Everyday AI
February 27, 2026
š Subscribe Here | š£ Hire Us To Speak | š¤ Partner with Us | š¤ Grow with GenAI
Sup yāall š
Aside from the upcoming deadline in the Anthropic-Pentagon fiasco, thereās been soooooo many big AI drops over the past day or three.
What are you most looking forward to using/learning about?
What recent AI update are you most looking forward to learning about and using?š³ļø Vote to see LIVE results š³ļø |
āļø
Jordan
Outsmart The Future
Today in Everyday AI
8 minute read
š Daily Podcast Episode: Three years ago, AI was a party trick. In Volume 10 of our Start Here series, we break down how it became an autonomous coworker. Give todayās show a watch/read/listen to find out.
šµļøāāļø Fresh Finds: Google adds āingredients to videoā, ChatGPTās Adult Mode Nears Release, Microsoft Rolls Out Copilot Tasks, and more Read on for Fresh Finds.
š Byte Sized Daily AI News: OpenAI Announces $110B Investment Push, 300+ Google and OpenAI Employees Back Anthropic in Open Letter, Meta Signs Multi-Billion Dollar TPU Deal With Google, and more Read on for Byte Sized News.
šŖ Leverage AI: How can you keep up with AIās massive glowup over the past 3ish years and actually know how to put it into practice? We gotchu. Keep reading for that!
ā©ļø Donāt miss out: Miss our last newsletter? We covered: Nano Banana 2 drops, AI leaders to meet with Trump, Perplexity launches safer āComputerā Agent, Claude Cowork launches scheduled tasks and more. Check it here!
Ep 723: From AI Chatbot to Autonomous Coworkers: How Consumer AI Has Changed and What's Next
If your entire company was using ChatGPT in 2022.... good chance you ended up in some trouble. š¬
Now if you're company is NOT using AI in 2026.... you'll definitely be in trouble.
But a different kind.
That's how quickly AI has changed the business landscape. What was once a megaviral party trick in 2022 is not the lifeblood of the American Enterprise.
So.... how the heck did it happen so fast?
In our 10th volume of the 'Start Here Series,' we break it all down.
Also on the pod today:
ā¢ AI chatbots: party trick origins š¤¹āāļø
ā¢ Scheduling: AIās hidden power ā°
ā¢ Autonomous coworkers on your desktop š»
Itāll be worth your 31 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 superset Orchestrates swarms of Claude Code, Codex, etc. in parallel, Mastra Code is a terminal-based AI coding agent built on Mastra's Harness, Alkemi is Data insights directly in Slack
Microsoft and OpenAI ā Original reports said Microsoft would be one of the three tech giants in OpenAIās latest funding round, but it was Softbank instead.
Xcode Update ā Xcode 26.3 adds autonomous coding agents that can read and modify your project. It supports OpenAI, Anthropic, and a permissions system to keep control.
ChatGPT Adult Mode Updates ā OpenAIās new ChatGPT Android update adds a āNaughty chatsā toggle that lets 18+ users request spicier, adult-themed language.
Google Vids Update ā Veo 3.1 converts short text into ready-to-post portrait videos. Admins can enforce custom templates for on-brand consistency.
Copilot Tasks ā Copilot Tasks turns chat into background work, automating tasks across apps. Join the research preview waitlist to try it.
Burger King Tests AI ā Burger King is adding an OpenAI-powered headset assistant named Patty to help staff with orders and maintenance.
OpenAI and PNNL ā OpenAI and PNNL tested coding agents on NEPA drafting and found they can cut 1ā5 hours per subsection, roughly a 15% drafting time drop.
1. OpenAI raises $110B to scale frontier AI with Amazon and NVIDIA backing š¤
In the largest-ever private funding round, OpenAI announced $110 billion in new investment at a $730 billion pre-money valuation, led by $30 billion each from SoftBank and NVIDIA and $50 billion from Amazon, plus strategic partnerships with Amazon and expanded inference and training capacity from NVIDIA.
The deal backs rapid global expansion of compute and distribution so OpenAI can push frontier models into everyday use, supporting surging product demand from 900 million weekly ChatGPT users and millions of business customers. The funding also boosts the OpenAI Foundationās stake value above $180 billion, increasing its ability to fund philanthropic efforts tied to health and AI resilience.
2. AI workers unite behind Anthropic as Pentagon pressure rises š
More than 350 Google and OpenAI employees signed an open letter in solidarity with Anthropic, urging their companies to refuse government requests to use advanced AI for domestic surveillance or fully autonomous weapons, a move made public within the last day.
The letter frames Pentagon negotiations with Google and OpenAI as an attempt to divide companies and pressures leadership to adopt unified red lines protecting public safety and civil liberties. Signatories include over 160 Google staff and 40 from OpenAI, and organizers say signatures are verified while allowing anonymity.
The Pentagon gave Anthropic until 5:01 pm ET today to allow the Pentagon to use its Claude AI model for "all lawful purposes" or lose its military contract and be treated as a supply-chain risk.
3. OpenAI links Codex straight into Figma for faster code-to-design workflows š§
OpenAI and Figma today announced a direct Codex-to-Figma integration that lets builders move seamlessly between code and the design canvas, accelerating iteration and delivery.
The connection uses Figmaās MCP Server to wire Codex into Figma Make, FigJam, and the platform so designs can be generated from code and implementation can flow back into code without losing context. The move tightens an already deep partnership that includes a Figma ChatGPT app and brings OpenAI models into Figmaās tools, aiming to blur role boundaries and speed product development.
4. Meta Leases Googleās TPUs in MultiāBillionāDollar AI Push šø
Meta has signed a multiāyear, multiābillionādollar deal to rent Googleās Tensor Processing Units, a timely move as companies race to scale AI infrastructure.
The agreement lets Meta develop advanced AI models on Googleās TPUs while it continues talks to buy TPUs for its own data centers next year. This adds to Metaās growing chip commitments, including large purchases from AMD and NVIDIA, and investors will watch how rising AI capital spending affects longāterm returns.
5.Block cuts more than 4,000 staff as AI reshapes work š¤
Block said Friday it will lay off more than 4,000 employees, almost half its 10,000-strong workforce, after its CEO, former Twitter Co-founder Jack Dorsey, warned that advances in AI are changing how companies must build and operate.
The move, announced alongside fourth-quarter results, pushed Block shares up as investors bet a smaller, AI-focused company can move faster despite the human cost. Dorsey predicted other firms will reach the same conclusion within a year, and his comments come amid roughly 30,000 tech layoffs so far this year as companies cite AI-driven efficiency changes.
Enterprises are straight up buying dedicated Mac Minis for their employees just so they can let AI agents run wild on dedicated desktops 24 hours a day.
But on the other end, we still have big businesses treating AI like a glorified search engine?
The gap between having a cute chatbot and a fully autonomous digital workforce is destroying corporate margins right now. Traditional ādigital transformationā and āAI councilsā sounds innovative and edgy in theory, but in reality, the red tape and meeting culture of corporate bureaucracy have caused a massive divide.
Leadership keeps shoving as many AI buzzwords as possible on their weekly all-hands, yet actual implementation still kinda looks like, āoh some employees chat with Copilot for better emails!ā
Lolz.
Hereās the reality: the growth of AI and Large Language Models has been unlike any technology weāve seen in the enterprise.
And if it feels like your company is trying to keep up but is still a year or so behind, thatās understandable.
Because while you were getting that random AI tool approved in 3 different committees, the AI landscape completely changed.
Thatās why we broke down AIās rapid timeline on today's Everyday AI and mapped out the exact evolution from basic AI chatbots to autonomous coworkers and more importantly, what comes next.
This is the real timeline weāve gotta be paying attention to. Letās learn.
1. Most Companies Are Three Phases Behind š„
When ChatGPT dropped in November 2022, most companies treated it like a fancy calculator and called it an AI strategy.
Phase two wired AI into company data, and enterprise teams bought their Copilot seats, ran a couple pilots, and declared themselves transformed.
They're still parked there today. Prolly putting it on their LinkedIn bio.
Phase three brought reasoning models like o1 preview and Claude 3.7 Sonnet that actually think before acting. Phase four gave you virtual AI coworkers like Manus and Genspark that take a goal and return finished work. Phase five is desktop agents like Codex and Claude Cowork running on your actual machine overnight, touching your files, your browser, everything local.
That jump from AI assistant to AI worker didn't take 10 years. It took 10 weeks, fam. The transition from late 2025 to early 2026 has seen crazy AI speed.
The executives who caught it aren't visionaries. They were just paying attention.
Try This
Run a five-minute audit Monday morning. Ask your team one honest question: are they using AI with actual reasoning capabilities, or just a glorified search engine wearing a chatbot costume?
Phase one and two looks like drafting emails and summarizing docs. Phase four and five looks like delegating a goal and coming back to finished work two hours later.
If the answer skews toward the first bucket, you've got a prioritization crisis dressed up as a tooling question. Most leadership teams are too scared to say that out loud. (And that's sorta why their competitors are already lapping them.)
2. Scheduling Is The Sleeper Feature ā”
In 48ish hours hours this week, Anthropic added scheduled tasks to Claude Cowork, Microsoft added them to Copilot, and Perplexity launched Computer, a 19-model system that runs entire workflows for hours or months in the background.
Google and OpenAI had this baked in for over a year already.
Most executives have genuinely never once clicked it. Wild, right?
(And prolly a little telling that your competitors have been quietly using it this whole time.)
Scheduling flips your AI from a tool you pick up to a worker you direct and forget about. The executives winning right now wake up every morning to finished competitive scans, briefings, and reports their AI ran overnight.
If yours ain't running while you sleep, you're leaving a full extra shift of output on the table every single night.
Try This
Open ChatGPT or Google Gemini right now and find "Tasks" or "Scheduled Actions." Most people have never clicked it even though it's been sitting there for months.
Pick one recurring deliverable your team dreads every week and schedule it. Let it run once and see what comes back.
If it's even 70% useful, that's hours back every week compounding hard over a quarter. This is prolly the highest-leverage 10 minutes you'll spend on AI this week, fam.
3. Rebuild Knowledge Work Then Pick Tools š
Most leadership teams are running internal Codex vs. Claude Code bake-offs while rebuilding how knowledge work actually gets done collects dust in nobody's calendar.
When every major platform has reasoning, scheduling, and desktop access, and they all do right now, the model is a commodity. The WORKFLOW is the differentiator.
AI agents already produce decks, docs, and spreadsheets that beat most human first drafts by default. The question ain't whether you'll use them. It's whether you rebuild processes around that now or duct-tape AI onto the same broken workflows you've always had.
Companies rebuilding right now are stacking compounding advantages. Companies still debating tools are gonna spend 2027 chasing a gap that keeps growing.
Try This
Pick one workflow this week that costs your team two to four hours and makes everyone groan. Write out every single step.
Then ask: what data does an AI agent need to run this end-to-end, what does finished output look like, and what decisions actually need a human?
Answer those three questions before buying another tool or running another training session. You don't need a new model. You need a new process. Start there before Deborah retires to Florida and takes all that institutional knowledge with her.
Reply