Building for a better planet and life. CPO & Co-Founder of Annwn - AI for sustainability at scale. https://teapawt.net/t.co/d46pENgrlp
London
Joined March 2010
- Tweets17K
- Following2.2K
- Followers1.1K
- Likes798
Pinned Tweet
“What is the use of living, if it be not to strive for noble causes and to make this muddled world a better place for those who will live in it after we are gone?”
6
Nvidia going red in more ways than one.
23
Adam Ball retweeted
Agents now make up the majority of web traffic, write more than half of new articles published online, and generate most of the code pushed to production.
This is what the early stage of a platform shift looks like: adoption is measurable, but the business models it enables mostly aren’t visible yet.
Welcome to the agent economy.
a16z.news/p/welcome-to-the-b…
Move over, B2B and B2C - the B2A economy is here
Agents are the primary users of the Internet, and soon will be the primary customers for every product
Our thesis @a16z on the massive opportunity here 👇 (cc @venturetwins)
If the token crunch hits, then this could be a win for those looking to keep lower levels of models off the frontier running. Market for old and new hardware goes up. Then you can have models running 24/7 for cheap by just paying for electricity.
Hook it up to a solar panel and you’d have sunshine powered, stardust based intelligence running round the clock.
Apple is working on a new app called Harbor
Harbor turns your Mac into a private AI inference server. Run open-weight models locally on Apple silicon—your data never leaves your hardware, and nothing reaches the cloud. Connect multiple Mac computers over Thunderbolt 5 to pool unified memory, enabling large models and higher concurrency that a single computer can handle alone.
Set up in minutes, not hours Harbor automatically detects your devices over Thunderbolt 5 to form a cluster. Plug in, start, and you’re running—no manual configuration, no command line, no infrastructure expertise required.
Download and run models directly Browse, download, and load open-weight models from within the app. Harbor handles model management so you can focus on the work, not the setup.
Cluster for scale Connect two or more Thunderbolt 5-capable Mac computers and Harbor pools their unified memory into a single inference resource. Run models that exceed any single computer’s capacity, or serve more concurrent requests across your cluster.
Harbor is currently in a private preview under the Claris brand.
31
When does the bazooka attachment become available?
Meet the Ring Drone Cam. A flying indoor security camera that delivers proactive wall-to-wall visibility, flying custom routes to the spots a fixed camera can't reach.
Check out the details: amazon.com/dp/B0H1PBKM3X
18
And just like that. Another one rolls off the tracks.
Gemini 4 Carbon is rolling out today. It’s built for sustained, long-horizon tasks, reasoning deeply through complex software engineering, cybersecurity, and scientific discovery.
Coming globally via the Gemini API in @GoogleAIStudio and Gemini for business, and for Ultra subscribers in @antigravity, the @GeminiApp, and AI Mode in Search.
18
Get me to the salmon farm rave
Norway now has more than 2,000 underwater AI laser robots shooting lice off its salmon
they're made by a company called Stingray, and they already guard more than 50 million salmon and trout 24/7
the reason is sea lice, tiny parasites that latch onto salmon and feed on them
as few as 11 can kill a small fish, and one pen can hold 150,000 salmon
the usual fix is pumping the fish out and blasting them with warm water or water jets to knock the lice off
but the handling is so rough that Norway's fish health report ranks it the top cause of death in full-grown farmed salmon
so instead, Stingray puts a robot inside the pen that hunts lice one at a time, all day:
> two cameras watch every salmon that swims past
> AI checks two frames to make sure it's really a louse
> then a green laser fires from up to 2 meters away
> the salmon's scales reflect the beam, so the fish swims off fine and the louse is dead
we now live in a world where AI lasers do pest control on your dinner lmao
24
This is incredible. AAA standard right there and only eight hours of Fable 5.2 cooking. Now to point this at helping accelerate enterprise sustainability teams problems at Annwn.
I wanted to test Fable 5.2's quality with one-shot prompts without human feedback.
Everything in this game is from scratch: models, sounds, textures. ~$550 at API pricing, ~2.7M output and 445M input tokens, heavily cached. Took 8 hours!
Play it here: claude.ai/artifact/BGY1SztrA…
18
Having novel data sets like this could be a huge unlock. Other frontier labs scraping the public we and buying off typical vendors vs proprietary ways of working and data like this to accelerate in novel ways.
Turns out that the Starlink software, built for managing the movement of vast amounts of data through space and to the ground, is also very well suited to training super intelligence
22
Feels like we are lowering our ambition in a number of places. We need to find that narrow corridor of decisions which mean we become more sustainable as a nation, but also one that thrives.
We should be bullish on EVs, renewable energy and storage. But also on AI development and its infrastructure. We can push on these and more if we’re truly ambitious and innovative.
A truly terrible decision at a truly terrible time. The shift to electric vehicles is government’s single biggest decarbonisation policy. Cutting EV targets would send a message on eve of #COP31 that UK is no longer serious about #climate action
theguardian.com/environment/…
22
We all need to care deeply about the things we build and put out into the world. For our Annwn customers I want their experience to be a joy when using our software. Not just deliver the outcome really well, but have fun doing it, be the place where you want to spend time during your working day.
There is still no substitute for caring. Every piece of software getting rapidly better is happening because a human cares. And happens to be using AI to move really fast.
22
This is fine as a statement but we are so far behind on energy, infrastructure and innovation. All the tech we’re using is broadly American.
UK will not slow down on AI despite backlash in US, says growth minister ft.trib.al/TtiOCoA
29
We’re finally seeing the move to proactive agents. Not waiting for you to initiate a turn. Having true agency. And then similar agents working to stop, learn, improve. Dreaming and connecting the dots on what’s gone on, what needs to improve.
Agents working around the clock. Working on your goals, improving over time and taking us to new places we hadn’t even imagined.
Things are going to get weird in a wonderful way.
1
24
It’s important to dream.
Have been thinking about this for Annwn. We take in enterprise data, then we parse it, structure it against our ontology and connect the dots. But what if those dots continue to grow and the connections beyond them after a first pass. Connecting up to other nodes and creating novel patterns. Going to be a huge unlock that we have AI reasoning and knitting data together.
Memory is what allows Claude to learn from experience. In a new Engineering Blog post, we share how “dreaming” helps Claude organize what it’s learned, correct its mistakes, and find what it needs more easily the next time. anthropic.com/research/compo…
30
