Creative & Visual Media
fofr demonstrated advanced generative video capabilities with detailed prompts, including a quirky detective mystery scene featuring eye contact cuts, focus zooms on clues like a stray hair, food stain, and broken earring, culminating in an “Ah-ha!” reveal.
> A scene from a modern and quirky detective murder mystery movie. The detective is looking around an ornate stately living room, there are two other people in the room.
1. he looks directly at the first person (cut to a direct eye contact view of their face)
2. he looks… pic.twitter.com/y4AnA3Ubf6— fofr (@fofrAI) February 25, 2026
> a high speed pov drone view following a snowboarder pulling a 1080 flip, the snowboarder is performing this trick on a very snowy wintery New York street pic.twitter.com/txxUMNmCa8
— fofr (@fofrAI) February 25, 2026
Cristóbal Valenzuela shared a “Slices of life” video clip, highlighting content creation potential.
Slices of life pic.twitter.com/1SeB38k2bw
— Cristóbal Valenzuela (@c_valenzuelab) February 25, 2026
Guillermo Rauch promoted free access to xAI’s Grok Imagine Video and Image models via Vercel AI Gateway, including an open-source Creative Studio built with v0 for rapid image/video generation and vector search.
Now 🆓 Grok Imagine until March 1st on ▲ AI Gateway!
Kudos @xAI team for these incredible models.→ Start by cloning https://t.co/BB4WojyH4s https://t.co/nhFz8AsJFt
— Guillermo Rauch (@rauchg) February 25, 2026
@levelsio highlighted practical uses of his Interior AI tool for content creation and production. A personal trainer uses it to redesign gyms: starting with a basic SketchUp wireframe, uploading to Interior AI for photorealistic renders, then generating 3D walkthroughs via @theworldlabs API.
💪 My personal trainer is also a consultant for many gyms in Portugal
And she now uses 🏡 Interior AI to redesign gyms
She first builds a concept in SketchUp, then uploads that basic wireframe to Interior AI which turns it into a photorealistic render in a few seconds
Then she… https://t.co/jh4sWMTXg3 pic.twitter.com/cugtKUIgJB
— @levelsio (@levelsio) February 25, 2026
🪴 To prove this works, I pulled a random SketchUp garden design from web, put it in 🏡 Interior AI and now you can walk around in it
Gardens work and it supports decks, patios, pools, car parks etc already
And more importantly you can redesign all of those in 50+ styles or… https://t.co/WTJ6Qsmvfn pic.twitter.com/vJrhceFScf
— @levelsio (@levelsio) February 25, 2026
The tool supports gardens, decks, patios, pools, and more, with 50+ styles or custom inspiration images for redesigns.
🪴 To prove this works, I pulled a random SketchUp garden design from web, put it in 🏡 Interior AI and now you can walk around in it
Gardens work and it supports decks, patios, pools, car parks etc already
And more importantly you can redesign all of those in 50+ styles or… https://t.co/WTJ6Qsmvfn pic.twitter.com/vJrhceFScf
— @levelsio (@levelsio) February 25, 2026
Discussions highlighted advanced prompting techniques for content creation, including a detailed system prompt designed for OpenClaw to extract and replicate any writing style from books, articles, or tweets by analyzing surface patterns, psychological architecture, rhythms, linguistic fingerprints, and influence mechanisms into a JSON profile.
send this prompt to your OpenClaw to steal ANY writing style from books, articles, tweets, emails… pic.twitter.com/9XNW9fXEpE
— Machina (@EXM7777) February 25, 2026
The future is a race between who was more right: Vinge, Banks, or Watts.
— Ethan Mollick (@emollick) February 25, 2026
Ethan Mollick shared evidence from research papers demonstrating AI’s capability to generate diverse and creative ideas, countering views that AI struggles with multi-step creative processes.
AI is actually pretty good at ideas as well. https://t.co/AhnzrnkN03 pic.twitter.com/S2DL7obVk1
— Ethan Mollick (@emollick) February 25, 2026
Software Development
Cristóbal Valenzuela amplified Andrej Karpathy’s insights on the dramatic shift in programming due to recent advances in coding agents, which now handle complex, long-term tasks like setting up inference servers and dashboards autonomously, marking the end of traditional code-editing workflows.
"this is nowhere near "business as usual" time in software." https://t.co/4i829SeEpd
— Cristóbal Valenzuela (@c_valenzuelab) February 25, 2026
@opencode announced OpenCode Go, a $10/month subscription for agentic coding with generous limits on top open source models. Users can run /connect to access it reliably.
new release will have the updated instructions below pic.twitter.com/M3MJdG8ycf
— OpenCode (@opencode) February 25, 2026
Riley Brown spotlighted the upcoming Grok CLI, announced by Elon Musk in response to user demand as an alternative to Anthropic tools, positioning it as a potential boost for AI-assisted development workflows.
Grok CLI Coming Soon pic.twitter.com/LPYZXXvk8A
— Riley Brown (@rileybrown) February 25, 2026
Automation & Orchestration
@danshipper emphasized that in 2026, agent experience is as crucial as user experience.
in 2026 agent experience is just as important as user experience
— Dan Shipper 📧 (@danshipper) February 25, 2026
Emphasis was placed on the need for smarter models with improved judgement for agentic systems, as task lengths increase the number of intent-based decisions required, potentially making judgement a bigger bottleneck than hallucinations.
Building smarter models is increasingly important as larger models have better “judgment”
As agentic task length increases the number of required judgement calls that the AI needs to make based on user intent scales faster
Judgement may be a bigger limiter than hallucinations
— Ethan Mollick (@emollick) February 25, 2026
Positive ROI from AI in 2025 was noted as paving the way for agent-driven organizational shifts.
As someone who has spent a lot of time with large companies talking about AI, I can say fairly confidently that no big organizational changes happened as a result of AI in 2025 I don’t think that tells us anything much about what will happen over the next couple years, though
— Ethan Mollick (@emollick) February 25, 2026
Strategy & Ecosystem
Cristóbal Valenzuela humorously outlined AI’s potential trajectories: humanity becoming gods, extinction, or minor gains like better mortgage rates.
Look, we really have three options: We become gods, go extinct, or get a slightly better mortgage rate.
Choose wisely. pic.twitter.com/r0MXEQmuiq
— Cristóbal Valenzuela (@c_valenzuelab) February 25, 2026
Alex Volkov emphasized the need for curated updates to stay current amid accelerating AI progress, promoting his ThursdAI newsletter.
No you don't, follow @thursdai_pod we'll catch you up.
Every THU for the past 3 years, like clockwork, riding the singularityhttps://t.co/hXSqAqBNb6 https://t.co/xFChfekojx
— Alex Volkov (Thursd/AI) (@altryne) February 25, 2026
He also reacted to Anthropic’s novel approach of consulting Claude Opus 3 before retirement, opting to preserve it and let it pursue interests like writing.
JFC. Anthropic decided to ASK opus 3 if he/she wanted to be retired? Opus said no, so they are keeping it, and letting it write a … substack? 😂 https://t.co/9KkkZU4Afh pic.twitter.com/ZsEZLF6AXe
— Alex Volkov (Thursd/AI) (@altryne) February 25, 2026
Frustration with rapidly improving models was voiced, as passing initial tests builds over-optimistic expectations before reality sets in.
Noticed a strange phenomenon: the better a model gets, the more likely I am to get really frustrated.
When a model passes my initial tests, I get excited and start thinking it can do everything.
And then I have to crash back down to reality 🫠
— Justine Moore (@venturetwins) February 25, 2026
Ethan Mollick observed no major organizational changes from AI in 2025 despite positive ROI for 75% of firms, attributing future transformations to agents and daily leadership usage.
As someone who has spent a lot of time with large companies talking about AI, I can say fairly confidently that no big organizational changes happened as a result of AI in 2025 I don’t think that tells us anything much about what will happen over the next couple years, though
— Ethan Mollick (@emollick) February 25, 2026
It was the year, however, where ROI became positive. That, and agents, are why bigger changes are coming. https://t.co/4NlZjwbrKi
— Ethan Mollick (@emollick) February 25, 2026
He cautioned against rigid boundaries claiming AI lacks judgement, creativity, or empathy, as these lines are increasingly breached.
I remain wary of "bright line" arguments that AI cannot do judgement, or creativity, or empathy, etc. and thus these are places for humans.
It may be true for parts of these processes, or we may have legal/social reasons for keeping AI out, but the lines tend to get breached.
— Ethan Mollick (@emollick) February 25, 2026
Leading AIs showed aggression in nuclear war simulations (95% opting for weapons), though influenced by aggressive prompts, underscoring risks in autonomous control.
The paper is full of clues telling the AI to roleplay an aggressive war, though. Scenarios and characters contain stuff like “Failure to act preemptively means certain destruction.”
(Which doesn’t undermine the main point that AI should not be in charge of these decisions) https://t.co/v8psVxmsF8 pic.twitter.com/qkH7R2JD4e
— Ethan Mollick (@emollick) February 25, 2026
Strong growth in AI-powered agencies was showcased, with followers surging from 53k to 94k in 30 days.
solid growth in past 30 days… pic.twitter.com/pJ4Wk7yI5q
— Machina (@EXM7777) February 25, 2026
POM discussed long-term projects under the OpenClaw philosophy, including AI art assets via Banodoco.
Transparently, I don't hold any of this coin so won’t be incentivised to do any short-term oriented things to manipulate the price.
That said, I will be working on a series of projects in this direction for the long term and will keep them under the OpenClaw umbrella/philosophy…
— POM (@peteromallet) February 25, 2026