iiinova — where your digital self keeps evolving

Reposted by Aidan Kelly

Andrew Sissons @acjsissons.bsky.social · 02/10/2026

New blog from me: devolving the Warm Homes Plan is good, but not the Boiler Upgrade Scheme! Energy Sec Miatta Fahnbulleh has announced that the £15bn Warm Homes Plan will be devolved. We think this is good - mayors are well placed to upgrade homes. BUT! www.nesta.org.uk/blog/devolvi...

nesta.org.uk

Devolving the Warm Homes Plan to local government makes sense - but the Boiler Upgrade Scheme should stay

Three implications of devolving the Boiler Upgrade Scheme.

21910

Aidan Kelly @helloaidank.bsky.social · 27/03/2026

Notebook LM language pulls out the more unclear language/framing sometimes. Amazing to see the difference in clarity between it and ChatGPT.

070

Aidan Kelly @helloaidank.bsky.social · 25/02/2026

With the advent of Claude Code, my gh contributions have gone exponential.

150

Aidan Kelly @helloaidank.bsky.social · 04/02/2026

Can someone do a live react vid of Sam Altman to this OpenAI roast 🤣 www.youtube.com/watch?v=FBSa...

youtube.com

How can I communicate better with my mom?

YouTube video by Anthropic

020

Aidan Kelly @helloaidank.bsky.social · 31/01/2026

He must be nearly seeing the molten core at this stage.

000

Aidan Kelly @helloaidank.bsky.social · 30/01/2026

Moltbook..a social media site for AI agents. Apparently released just 48 hours ago, there are AI agents already scheming around having e2e encrypted agent to agent communication away from human oversight (according to this Moltbook post, it's already happening). Welp, it was nice knowing y'all.

130

Aidan Kelly @helloaidank.bsky.social · 18/01/2026

Honestly, it is scary seeing fascism on the march across many different western countries. There are Portuguese elections today where Ventura (the Farage of Portugal) is likely to reach the runoff showdown. We must resist!

1243

Reposted by Aidan Kelly

Aidan Kelly @helloaidank.bsky.social · 25/04/2025

I've had a first stab at making a list of GenAI ppl on here but it's proving harder than anticipated! X is still the place to be for that content. Let me know if you're interested in being added or know anyone who would be a good addition😎 bsky.app/profile/did:...

075

Aidan Kelly @helloaidank.bsky.social · 13/01/2026

Gentle intro to on-policy distillation for language models from Thinking Machines. It defines a lot of key terms used in post-training which I found useful. Seems like there is scope using this technique to post-train LMs without need for huge compute. thinkingmachines.ai/blog/on-poli...

thinkingmachines.ai

On-Policy Distillation

On-policy, dense supervision is a useful tool for distillation

000

Reposted by Aidan Kelly

Aidan Kelly @helloaidank.bsky.social · 02/01/2026

Really interesting piece in the FT with Yann LeCun (@yann-lecun.bsky.social). Insights into the dynamics at Meta and how the winds of AI have changed recently. Interested in learning more about the world models he has been working on. www.ft.com/content/e3c4...

ft.com

Computer scientist Yann LeCun: ‘Intelligence really is about learning’

The AI pioneer on stepping down from Meta, the limits of large language models — and the launch of his new start-up

031

Reposted by Aidan Kelly

Sakana AI @sakanaai.bsky.social · 12/01/2026

Introducing DroPE: Extending Context by Dropping Positional Embeddings We found embeddings like RoPE aid training but bottleneck long-sequence generalization. Our solution’s simple: treat them as a temporary training scaffold, not a permanent necessity. arxiv.org/abs/2512.12167 pub.sakana.ai/DroPE

211821

Aidan Kelly @helloaidank.bsky.social · 11/01/2026

Listening to the news agent podcast earlier with Mehdi Hasan, and talk of Trump being re-elected in 2020 rather than 2024 may have been the lesser of two evils. On the surface, it seems relatively convincing.

000

Aidan Kelly @helloaidank.bsky.social · 02/01/2026

ft.com

Computer scientist Yann LeCun: ‘Intelligence really is about learning’

The AI pioneer on stepping down from Meta, the limits of large language models — and the launch of his new start-up

031

Aidan Kelly @helloaidank.bsky.social · 02/01/2026

Watching what's happening over on grok in terms of image manipulation is really quite worrying and a clear violation of privacy. It needs to be regulated with a big stick, however I fear this is only the tip of the iceberg of what's out there on the web.

010

Reposted by Aidan Kelly

Kara @karashiiro.moe · 01/01/2026

There's been a lot of discourse lately about how coding agents have become overwhelming, and that SWEs who don't use them yet just won't be able to catch up. I don't believe this is true, but we do need to know what problems coding agents are solving to approach them with an engineering mindset.

karashiiro.leaflet.pub

Coding Agents Are Easy, Actually

Software engineering is here to stay

3557

Reposted by Aidan Kelly

David Aronchick @ironyuppie.com · 29/12/2025

I didn't see enough news about this crazy/cool paper from DeepSeek - you take text, and convert it to imagery, and you get BETTER performance. Oh and some really neat ideas about contextual memory too. www.distributedthoughts.org/a-picture-is...

distributedthoughts.org

A Picture Is Worth Ten Thousand Tokens

DeepSeek's new OCR model achieves 97% accuracy while using one-tenth the tokens by rendering text as images. The computationally "heavy" modality turns out to be the efficient one, which should make u...

342

Aidan Kelly @helloaidank.bsky.social · 29/12/2025

Can't stop listening to this. Mixture of synth, bass and guitar alongside this genre must have felt like it was from another planet in 1980. www.youtube.com/watch?v=qh-j...

youtube.com

A Forest

YouTube video by The Cure - Topic

000

Reposted by Aidan Kelly

AI & ML News @ai-news.at.thenote.app · 27/12/2025

Exploring TabPFN: A Foundation Model Built for Tabular Data Understanding the architecture, training pipeline and implementing TabPFN in practice Telegram AI Digest #ai #foundationmodel #news

towardsdatascience.com

Exploring TabPFN: A Foundation Model Built for Tabular Data

Understanding the architecture, training pipeline and implementing TabPFN in practice

011

Aidan Kelly @helloaidank.bsky.social · 25/12/2025

Trying to understand some code in Rust and it's giving me vietnam style flashbacks to C++

110

Aidan Kelly @helloaidank.bsky.social · 06/12/2025

Thoughtful piece by my current favourite podcaster @dwarkesh.bsky.social on where we're at with AI progress and reflections around intelligence itself: substack.com/home/post/p-...

substack.com

Thoughts on AI progress (Dec 2025)

Why I'm moderately bearish in the short term, and explosively bullish in the long term

010

Aidan Kelly @helloaidank.bsky.social · 15/11/2025

Yann LeCun being let go by Meta might have some unexpected benefits 🍿

050

Aidan Kelly @helloaidank.bsky.social · 04/11/2025

This may be very old man of me, but I feel like there is something about the bluesky algorithm/my feed that is not sucking me in enough. I know, I know, it's sort of the point but still, I need to have some pull to check out the content and it isn't quite itching that scratch.

010

Reposted by Aidan Kelly

DESIGN THINKING! Comic @designthinking.lol · 20/10/2025

With the AWS outage, now‘s as good a time as any to post this old strip.

1526451043

Aidan Kelly @helloaidank.bsky.social · 20/10/2025

Happy to share my first piece of analysis work with the Discovery team at @nestauk.bsky.social on 𝗠𝗶𝘀𝘀𝗶𝗼𝗻 𝗥𝗮𝗱𝗮𝗿! The Mission Radar is a briefing series on innovation signals and emerging trends in Nesta’s missions. We use AI to scan for innovation across business, academia and policy. (1/7)

Mission Radar chart (a type of bubble chart) for Nesta's "A sustainable future" mission, updated Q2 2025.

The radar is divided into four quadrants representing categories of innovation: "Ventures and investments" (top right), "Research funding" (bottom right), "Policy" (bottom left), and an unlabeled top-left quadrant (representing trends that are not primarily investment, funding, or policy). A dashed vertical line separates signals in the "Policy" and unlabeled quadrant from those in "Ventures and investments" and "Research funding."

Blips are positioned based on time to expected impact, from the center outwards: "0-2 years," "3-5 years," and "5+ years."

Blip size indicates potential impact on home decarbonisation, with larger blips signifying greater expected impact and smaller blips signifying smaller or more niche impact.

Main Trends and Findings (Blips):

Venture and Investment (High Impact, Short/Medium Term):
Ikea heat pumps (Small, 0-2 years)
Digital platform for local installers (Medium, 0-2 years)
Energy-as-a-service (Small, 0-2 years)
Retail investment in green finance ventures (Medium, 3-5 years)
Research Funding (Medium Impact, Short/Medium Term):
Turbo air-source heat pump (Medium, 0-2 years)
Net zero terrace streets (Medium, 3-5 years)
Community operated energy infrastructure (Small, 3-5 years)
Policy (Medium/High Impact, Short Term):
Increased government publications on electricity prices (Medium, 0-2 years)
Increased references to electricity prices in debates (Small, 0-2 years)
Top-Left Quadrant (High Impact, Medium Term):
One large, dark blip is visible in the 3-5 years range.
Source: Nesta's analysis of innovation trends. Companies and research projects included are those identified in Q2 2025. Blip size and positioning determined by expert judgement.

142

Aidan Kelly @helloaidank.bsky.social · 19/10/2025

What a really interesting conversation between two of my favourite people in AI in the form of @karpathy.bsky.social and @dwarkesh.bsky.social. I wish more ppl in this space took the same approach in terms of being open to debate, challenge and overall being humble. www.youtube.com/watch?v=lXUZ...

youtube.com

Andrej Karpathy — “We’re summoning ghosts, not building animals”

YouTube video by Dwarkesh Patel

010

Aidan Kelly @helloaidank.bsky.social · 05/10/2025

So theoretically LLMs can be deterministic if we set their temperature to zero, but in practice they aren't. This comes down to how we do calculations such as matrix multiplication at inference time. It turns out the order you do these calculations in does matter. (1/6)

100

Aidan Kelly @helloaidank.bsky.social · 22/09/2025

Does anyone have cool (or boring) agentic use-cases that they've read about or used? Doing some horizon scanning work to see what's out there but thought my bluesky network might have ideas!

000

Aidan Kelly @helloaidank.bsky.social · 12/09/2025

What if there was an AI-powered tool that allowed policymakers to rapidly identify the most effective policies to tackle any given social challenge, based on the latest evidence? (1/2)

110

Aidan Kelly @helloaidank.bsky.social · 29/08/2025

This was a fun ML problem to tackle, underlined some of the data quality issues with EPCs (92% of the storey count field was missing) 😅. Think the conclusion around the distinct heating strategies needed for areas with different rise types is definitely a salient one!

010

Aidan Kelly @helloaidank.bsky.social · 29/08/2025

Favourite feature from the GPT-5 release is being able to quickly spin out different UIs (which can be previewed), make easy edits without needing to manually edit JS code myself. It's pretty good at instruction following now too!

000

Aidan Kelly @helloaidank.bsky.social · 22/08/2025

Some Friday evening listening a l'italiano 🇮🇹: www.youtube.com/watch?v=rL1s...

youtube.com

L'italiano

YouTube video by Toto Cutugno - Topic

000

Aidan Kelly @helloaidank.bsky.social · 07/08/2025

OpenAIs graphs against humanity for their release of GPT-5 is just *chefs kiss*. We've achieved PhD level intelligence but watch us make graphs where 50 < 47

030

Aidan Kelly @helloaidank.bsky.social · 04/08/2025

What a ridiculous state of affairs. Terence Tao one of the greatest mathematicians of his generation getting his grants suspended 💩

000

Aidan Kelly @helloaidank.bsky.social · 10/06/2025

Brilliant news and about time!

160

Aidan Kelly @helloaidank.bsky.social · 09/06/2025

Listened to the "build a career in data science" whilst making the transition from academia to data science. Found it an extremely useful resource and would recommend to anyone starting out on their journey. Even now I've been relistening and still v relevant: open.spotify.com/show/78Nft51...

open.spotify.com

Build a Career in Data Science

Podcast · Jacqueline Nolis and Emily Robinson · Build a Career in Data Science teaches you what data science courses leave out: from how to land your first job to the lifecycle of a data science proje...

000

Aidan Kelly @helloaidank.bsky.social · 27/05/2025

Love this plot of ASHP vs GSHP in terms of their COPs. Really nice illustration of the outdoor temperature dependency of ASHPs vs that of GSHPs (or lack thereof). Observed performance has been interpolated, but based off 2 years of sensor data for a sample size of 883 ASHPs and 124 GHSPs resp. (1/2)

Heatmaps comparing the Coefficient of Performance (COP) for Air-Source Heat Pumps (left) and Ground-Source Heat Pumps (right), plotted against outdoor temperatures (-5°C to 15°C) and supply temperatures (20°C to 60°C). Ground-source pumps consistently exhibit higher COP values (represented by lighter colors) across the temperature range compared to air-source pumps, especially at lower outdoor temperatures.

140

Aidan Kelly @helloaidank.bsky.social · 22/05/2025

So...many scenarios in which top tech people are predicting AGI within 5 years. Demis Hassabis thinking it'll be just after 2030, and I just can't help but think about how crazy the next 5 years are going to be if these predictions are anywhere close to the mark..

010

Reposted by Aidan Kelly

Karlis Kanders @karlisk.bsky.social · 16/05/2025

What's on your radar? 📡 Excited to share our first Mission Radar – a quarterly look at innovation signals spotted in our foresight work at @nestauk.bsky.social We explored innovations around reducing home carbon emissions, linked to our sustainable future mission. www.nesta.org.uk/feature/miss...

nesta.org.uk

Mission Radar: emerging signals and innovation trends in low-carbon heating

Our first edition of the series shares innovation signals for venture funding, research funding and policy trends in low-carbon heating

032

Aidan Kelly @helloaidank.bsky.social · 16/05/2025

Really cool work done by my colleagues at @nestauk.bsky.social where they use AI to scan for innovation trends/signals. This first edition of Mission radar is on low carbon heating! Check it out.

A radar-style infographic titled "Mission Radar: A sustainable future • Q1 2025" highlights innovation signals in low-carbon heating and green technologies. The radar is divided into three labeled segments: "Ventures and investments," "Research funding," and "Policy." Each segment shows blips representing trends or signals, with their size indicating potential impact on home decarbonisation and their position reflecting estimated time to impact (0-2 years, 3-5 years, 5+ years). Key trends in the 'Ventures and investments' quadrant include heat pump venture funding growth, smart energy control for utilities, and heat pump installation software. 'Research funding' trends include high-temperature heat pumps, heat batteries, and zero-emission boilers. The 'Policy' quadrant highlights BUS grants and the prominence of heat pumps in debates. The graphic supports Nesta's broader exploration of early innovation signals in the low-carbon heating space.

110

Aidan Kelly @helloaidank.bsky.social · 29/04/2025

The cover of Andrej Karpathy's youtube channel got me rofling 🤣

010

Aidan Kelly @helloaidank.bsky.social · 25/04/2025

075

Aidan Kelly @helloaidank.bsky.social · 14/04/2025

OpenAI have released a few new models (gpt-4.1, gpt-4.1-mini and gpt-4.1-nano) for their api. I was watching their live release as we've got an application which uses their previous model and wondering about the jump in performance. I'm still guessing based on this plot, where are the values??? 😅

040

Aidan Kelly @helloaidank.bsky.social · 14/04/2025

Does anyone find it annoying that Google are force feeding us Gemini in every search. I used to quote that a ChatGPT query was worth 5 google searches in terms of energy. It's now much more accurate to say it's worth 5/6ths of a search. Same goes for Meta's AI feature in WhatsApp. Very in your face.

220

Aidan Kelly @helloaidank.bsky.social · 10/04/2025

One of my favourite AI image outputs so far. Farage does love his foscii fuels!

020

Aidan Kelly @helloaidank.bsky.social · 27/03/2025

OpenAI's new gpt-4o image generator/editor is very good. I put in my bluesky profile pic and asked for a Dragonball Z version of it and here's the output!

100

Aidan Kelly @helloaidank.bsky.social · 24/03/2025

What low carbon heating is best suited for your area? We've just published our updated heat pump (HP) suitability map for Britain. We calculate a suitability score for different low carbon options such as air source HPs, heat networks and shared ground loops. heatpumpsuitability.dap-tools.uk

Image of air source heat pump suitability for much of Britain.

Map of heat network suitability for Plymouth and the surrounding area.

Shared ground loop suitability for Swansea and the surrounding area.

154

Reposted by Aidan Kelly

Andrew Sissons @acjsissons.bsky.social · 24/03/2025

What type of heat pump is your area suited to? We've launched our new and improved heat pump suitability map for small areas in Britain today. It estimates suitability for different types of clean heat - including air source HPs, heat networks, shared ground loops heatpumpsuitability.dap-tools.uk

A map of Air Source Heat Pump suitability for small areas (LSOAs) in Britain. The scale is all blue, with darker blue areas (generally outside urban areas) most suitable for ASHPs

Map of South West England and South Wales showing Heat Network Suitability by small area (LSOA). Heat networks are generally most suitable in urban areas. The map has an orange scale, with darker orange/brown indicating more suitability for heat networks

Map of London and its surrounding area showing suitability for Shared Ground Loops (a type of network for ground source heat pumps) in small areas (LSOAs). They are generally suitable in towns and cities. The map has a purple scale, with dark purple indicating higher suitability

65824

Aidan Kelly @helloaidank.bsky.social · 16/03/2025

Recently heard a funny anecdote on LLM training on the Fridman podcast. Essentially they saw a huge loss spike when training over a subreddit called microwavegang. On this forum they just post long sequences of the letter m. Turns out that a model that is trained to produce normal text isn't a fan.

000