Tech Landscape #435
Ads & creators at Cannes Lions, more agents in creative tools, and broadcast-ready video models
Hello!
Instead of an intro I’ve written about the workflow I use to write this email, which you can find at the end if you’re interested.
So, let’s get on with it. Hope you’re well!
Synthetic Content
Lots of small quality-of-life improvements to video models this month: Happy Horse 1.1 adds “enhanced motion, texture, and consistency”, Kling O3 was upgraded with stronger prompt and reference consistency (apparently; I can’t find an official source for this), and Seedance 2.0 can now generate in native 4K (at ~$1.50 per second!).
Here ⬇️ are a couple of quick tests of Seedance 2.0 in 4K (best viewed in 4K, obvs.); the detail is exceptional.
Seedance 2.5 was announced, with 30-second generation from up to 50 reference images, set to be launched in July. ark.volcengine.com
I’ll have much more on this when it’s released.Krea released the open weights of its Krea 2 model, making it available to developers for use in custom workflows. instagram.com/krea_ai
Spaitial launched a Claude plugin and MCP server, enabling developers to build 3D environments through coding interfaces. x.com/SpAItial_AI
Creative Tools
Adobe is buying Topaz Labs, the company that specialises in video and image enhancement and upscaling. news.adobe.com
Google Flow’s Agent can use Google Maps Street View for real-world location grounding. instagram.com/googleflow
It works globally, but only for US locations. You can see it in my example ⬇️, with the Street View tile inset. This is powered by the computer use capability newly arrived in Gemini 3.5 Flash (see below).
OpenArt launched Director, an AI video generation assistant that turns conversational prompts into short films (up to five minutes long), maintaining visual consistency across characters, scenes, and brand assets. openart.ai
They’re trying to call this “vibe directing”; please, no…
Runway’s Agent 2.0 is an upgraded conversational AI assistant designed to help marketers automatically generate tailored ads, videos, and social media campaigns, and analyse performance metrics. runwayml.com
Runway can now localise ads, translating any static image into many languages. instagram.com/runwayapp
ElevenLabs’ ElevenCreative launched Ads Engine, that connects directly to Google, Meta, and LinkedIn ad platforms, initially to localise ads by translating text, adapting images, and dubbing videos, with more capabilities to follow. elevenlabs.io
Luma Agent added Connectors to third-party services, starting with Airtable, Dropbox, and Google Drive. instagram.com/lumalabsai
Design & Dev
Figma announced several product updates at its Config 2026 event. The headline is code layers, which makes every layer on the canvas editable with code, unifying design and build in a single interface; it’s coming soon in closed beta. Figma Motion is a timeline editor for the canvas, enabling motion design of every element. Figma and Weave are becoming more tightly integrated. And the design agent is coming to all users, and now lets them code their own shaders and plugins.
figma.com/blog/config-2026-recap/
As most of the big creative / design tools are completely reinventing themselves with AI, it’s going to be harder and harder for the refuseniks to opt out.
Google’s AI Studio added Design Variations, which generates quick previews of visual designs that can be applied to apps. x.com/GoogleAIStudio
Genspark launched Design, for agentic creation of UI prototypes, videos, HTML animations, posters and more. instagram.com/genspark_ai
Assistants
OpenAI announced the GPT-5.6 model family: Sol is the flagship, built for advanced reasoning and cybersecurity tasks; Terra is “a balanced model for everyday work”; and Luna is the faster, cheaper variant. They will initially be available as a limited preview to trusted testers, before a broader rollout. openai.com
The delayed release is because of US government oversight; expect this to become the norm, at least for the immediate future.
Google’s Gemini 3.5 Flash added computer use, letting developers build AI agents that can see, reason, and take action across browser, mobile, and desktop environments. blog.google
Claude launched Tag, a new way for teams to work together with the assistant in collaboration platforms, starting with Slack. anthropic.com
The Poke assistant can ask humans to fulfil tasks, using gig economy marketplaces. x.com/interaction
If there’s a task it can’t complete without human interaction, it can hire one. This is pretty wild. People talk about AI with “a human in the loop” but I don’t think they mean this. It’s available as a Preview for Ultra ($200pcm) subscribers.
Social
Instagram is expanding its TV app, adding interest-based channels, Stories, casting from mobile, and more, and coming to Samsung Smart TVs (in the US). about.instagram.com
YouTube updated its Shorts player, adding a distraction-free Clear Screen mode, playback speed controls, a heart icon to replace the thumbs-up, and retiring the dislike button in favour of more precise feed controls. blog.youtube
Cannes: Ads & Creators
The Cannes Lions festival offers an opportunity for the big platforms to announce their new advertising and creator tools, either at the event itself or riding its PR wave.
Meta unveiled an end-to-end creative solution for brands that can handle strategy, creative and production (with brand memory), distribution, and testing. Plus there are AI upgrades to Ads Manager, and a unified Meta Creator Marketing Hub to streamline influencer campaigns. facebook.com
The company was at pains to state that it’s not in competition with agencies, but it is.Facebook is bringing back Creator Studio, now reimagined as an AI-powered companion app with advice on how to grow an audience. creators.facebook.com
YouTube announced Gemini-powered insights tools to help brands and agencies optimise creator marketing. blog.google
Reddit introduced new advertising tools, powered by Reddit Community Intelligence, including an AI-driven free-form ad generator, tailored creative assets, Redditor Highlights, and multi-advertiser Shopping Listing Ads. redditinc.com
TikTok added Symphony Agent to assist with the creation of ads in the Symphony Creative Studio. ads.tiktok.com
Amazon introduced Alexa+ Agentic Ads, an ad format for Echo Show devices that lets users complete purchases entirely within a conversation. advertising.amazon.com
Everything Else
Meta introduced Meta Glasses, its first line of smart glasses to carry only its own name, with the same features as the Meta Ray-Ban series (including the new Muse Spark model) but at a lower price point. about.fb.com
Google Play now allows more choice of payment providers, and has restructured its fees. android-developers.googleblog.com
This is a global change driven by European legislation, and is coming to the UK and EEA first.
My Workflow
The increased pace of change in the last few years means I have a lot more to cover in this newsletter than I did when I started, and it takes up more of my time. Recently I’ve been looking at ways to use automation and AI to help me with some of that burden. But it’s a balancing act; I don’t want to automate too much, because then it loses any value. So here’s my current workflow:
Throughout the week, I save all the links in Airtable
At the end of the week I paste all of the links into a custom app I built, which uses Gemini to summarise them in an approximation of my voice.
Some links (especially social links) are blocked to crawlers, so I write those myself
When all the summaries are generated or written, I paste them back into Airtable
I export the Airtable (as a CSV file) and import it to another custom app I wrote which formats it for this newsletter
I post that into Substack, then do a few rounds of editing, writing, and rewriting
Those final rounds are really important; the automated summaries are very rarely perfect, I almost always have to make tweaks or changes. They could be to make a sentence more human, to change American spellings, or to focus more on the things that I think are important.
For example, this is how Gemini summarised one of this week’s stories:
Adobe announced a definitive agreement to acquire Topaz Labs, an AI company specializing in advanced video and image enhancement models.
And this is what I edited it to:
Adobe is buying Topaz Labs, the company that specialises in video and image enhancement and upscaling.
Whereas this one I had to rewrite entirely; not because it’s wrong, but because I think it focuses on the wrong features:
Figma summarized new product updates from Config 2026, an update that introduces a built-in timeline animation feature, AI-generated shaders, and prompt-driven custom plugins.
And I write all of the comments and asides myself. I reckon this process saves me an hour or more each week, but maybe 75% is still hand-written. My current big project is working out ways to make the process much simpler, while bringing that writing ratio down to about 50%, but still preserving my voice.