Fintan Ryan

Competitive Insights at @github, much #AI, but still tracking #DevSecOps
& #Cloud l recovering analyst (ex @redmonk, @Gartner_Inc) | 331 ppm

[Fintan Ryan @fintanr]

We are all learning a lot about the ethics and values of various companies right now. Keep notes.

[Erin Havens @erinhavens]

@GHchangelog https://t.co/7s1stuQKqS

[Fintan Ryan @fintanr]

This 👇

[Darren Shepherd @ibuildthecloud]

Learning to code is not worthwhile. Learning computer science is. Return to the fundamental. Understand the systems, command the systems, don't be a dumb operator. For too long developers have constrained themselves to mastering a programming language and ecosystem. If you take a step back and understand the fundamentals, you can widen your scope to more languages and more ecosystems because the similarities exist but only the syntax differs.

[GitHub Changelog @GHchangelog]

Kimi K2.7 Code, an open-weight model, is now generally available as a selectable option in GitHub Copilot.

• Available on Pro, Pro+, Max plans with rollout starting in Visual Studio Code and other dev environments

https://t.co/ESduzvz7TO

[Martin Woodward @martinwoodward]

Anyway, folks have been asking for more data on this so we published our internal results here: https://t.co/dhCMzAAhbC

Let me know what you think. Is this type of data interesting to you? How does this compare with your real-world use?

[Fintan Ryan @fintanr]

I recall in Sun back in the early 2000s that timezones were an oddly complex problem for some people on the US west coast to grok. Funny how certain things are constants in tech once you move away from those that know how to work async.

[Fintan Ryan @fintanr]

Congrats to the @ona_hq team, been a long journey from their GitPod days, but thats a solid ending for this stage and an exciting start for the next.

[OpenAI Newsroom @OpenAINewsroom]

We’ve reached an agreement to acquire @ona_hq.

Its secure cloud execution technology will help Codex take on longer-running work, even when laptops are closed, and help more organizations deploy agents securely in production.

After closing, Ona will join OpenAI’s Codex team. https://t.co/RXoL4LSDcf

[Fintan Ryan @fintanr]

Lunch is a Legacy System: Deprecating Human Needs for Q4 Targets (satire) - https://t.co/Wxil22HScM

[Fintan Ryan @fintanr]

Some quick thoughts on @Cloudflare 's acquisition of @voidzerodev and various recent developer tool acquisitions by @Anthropic, @OpenAI , @vercel and @Cloudflare.

https://t.co/Is4hWf38OL

[Fintan Ryan @fintanr]

well thats a new one, signing up for a newsletter which wants to share my e-mail with one of their sponsors to be signed up to their e-mail list before I can fully subscribe. I guess its more honest than just sharing the info, but its a no from me.

[OpenUK @openuk_uk]

One week to go to State of Open Con on the Road, Edinburgh. The perfect opportunity to learn about AI and open source from world-leading experts. @AmandaBrockUK, @adrianjhpc, @adrianmouat, Anastasia Stasenko, @sublimino, Camilla de Coverly Veale, Chaamini Mangaleswaran, Chris Plank, @clry2, Dave Buckley, @fintanr, Iain G. Mitchell KC, Jean-Baptiste Kempf, @justincormack, Juliette Denny, Keith Bergelt, Lee Fulmer, Liz Rice, Mallory Durran, Matt Jarvis, Matt Barker, @MikeMcQuaid, Nick Jones, @PaulaLKennedy, Sachiko Muto, @KimmichSal, and Tom Wilkinson.

We're also joined by the table holders: @Percona, @NatWestGroup, The National Robotarium by @HeriotWattUni, Occasio, @growthengineer, and @cybersecure_in.

See the speakers, schedule and get tickets at https://t.co/m2A74XvxB6

#opensource #AI #opensourceai #aiskills #stateofopencon #soocon26 #openuk #openhq #aiscotland #AI #aimodels #aiagents

[Fintan Ryan @fintanr]

The old Mark Twain adage of “I didn't have time to write a short letter, so I wrote a long one instead” has never been more true.

[Fintan Ryan @fintanr]

This. The social contract of actually writing and understanding what you have written is vital.

I have read, and discounted until I talk to the owner, *so many* reports & docs recently, purporting to be from a human, that are largely generated by LLMs.

[Marc Brooker @MarcJBrooker]

Great question.

I don't use LLMs for writing.

I use agents extensively for brainstorming, research, checking facts, handling markup, finding references, indexing data, and so on.

But I think that asking people to read LLM-generated text breaks a kind of social contract.

[Marc Brooker @MarcJBrooker]

Great question.

I don't use LLMs for writing.

I use agents extensively for brainstorming, research, checking facts, handling markup, finding references, indexing data, and so on.

But I think that asking people to read LLM-generated text breaks a kind of social contract.

[Simon Willison @simonw]

The conclusions here feel wrong to me. The two lessons I see are:

1. Don't run agents anywhere they might be able to access production environment credentials - it's on you to know which credentials those are

2. Keep tested backups that are independent from your production host

[Samay @Samaytwt]

Unpopular opinion: "AI makes everyone a developer" is true the same way "cameras makes everyone a photographer"

[Fintan Ryan @fintanr]

A little fun for a Monday, I watch a lot of rugby*, and like anyone following the @URCOfficial there are lots of possible permutations for the final three rounds, so I put together a little app to stick in the results you expect and see what the table might look like https://t.co/Y19AHlI5JC

[Satya Nadella @satyanadella]

In GitHub Copilot CLI, you can now leverage a multi-model reflection loop as a reviewer.

Super helpful for catching issues early before they compound.

[Fintan Ryan @fintanr]

@gabemonroy Congrats @gabemonroy, lots of interesting challenges ahead for you.

[Evan Boyle @_Evan_Boyle]

Introducing RubberDuck in Copilot CLI: a new builtin subagent for cross model family escalations. Sonnet will now proactively solicit input from GPT 5.4 and vice versa.

In our most difficult benchmark subsets, this results in a massive 5% improvement in resolution rates

[Martin Woodward @martinwoodward]

Unlimited API tokens across ALL the coding models via GitHub Copilot is low key one of the best perks for engineers working at Microsoft.