Liz Keogh
@lunivore.bsky.social
Staff Developer, Lean/Agile coach, speaker, blogger, pattern hunter, poet, writer, complexity thinker, free will skeptic, Londoner, Goth, she/her.
Tried to find the page documenting my main No Man's Sky base. I do also play Palworld. This is hilarious.
May 10, 2025 at 5:13 PM
Tried to find the page documenting my main No Man's Sky base. I do also play Palworld. This is hilarious.
Software devs solve problems that haven’t been solved before, or solve them in new contexts; otherwise we’d be buying (or open sourcing) solutions.
When AI solves problems that don’t come from its training data set, I’ll worry about my mortgage.
Until then it’s great for the other stuff.
Mostly.
When AI solves problems that don’t come from its training data set, I’ll worry about my mortgage.
Until then it’s great for the other stuff.
Mostly.
May 2, 2025 at 10:51 PM
Software devs solve problems that haven’t been solved before, or solve them in new contexts; otherwise we’d be buying (or open sourcing) solutions.
When AI solves problems that don’t come from its training data set, I’ll worry about my mortgage.
Until then it’s great for the other stuff.
Mostly.
When AI solves problems that don’t come from its training data set, I’ll worry about my mortgage.
Until then it’s great for the other stuff.
Mostly.
AI is super-helpful, especially when we tread on someone else's specialist turf to get things done. But it still takes a human to help you check / test the work, and AIs can't tell you who the right person is to do that, or reach out to them for you. As ever, the hardest problems in SW are human.
April 18, 2025 at 10:31 AM
AI is super-helpful, especially when we tread on someone else's specialist turf to get things done. But it still takes a human to help you check / test the work, and AIs can't tell you who the right person is to do that, or reach out to them for you. As ever, the hardest problems in SW are human.
Specialisms in modern SW means that we have to work with each other: to over-communicate, relentlessly clarify priorities, work as a team-of-teams when they shift, create and clean up docs, and it's always messy no matter how hard you try. Katherine Kirk calls it "the entanglement of entanglement."
April 18, 2025 at 10:27 AM
Specialisms in modern SW means that we have to work with each other: to over-communicate, relentlessly clarify priorities, work as a team-of-teams when they shift, create and clean up docs, and it's always messy no matter how hard you try. Katherine Kirk calls it "the entanglement of entanglement."
It's been eye-opening getting "back to the coal face". The complexity of modern software development is mind-boggling. Back in 2000 one person could reasonably hold the whole of an Enterprise website plus services in their head having spent a year or so on it. Now we necessarily have specialisms.
April 18, 2025 at 10:23 AM
It's been eye-opening getting "back to the coal face". The complexity of modern software development is mind-boggling. Back in 2000 one person could reasonably hold the whole of an Enterprise website plus services in their head having spent a year or so on it. Now we necessarily have specialisms.
Reposted by Liz Keogh
I stumbled across this essay as I was falling asleep last night, and left myself multiple emphatic reminders to remember to circle back to it today.
"The joy we're about to lose...we'll become overseers rather than creators, managers rather than builders."
annievella.com/posts/the-so...
"The joy we're about to lose...we'll become overseers rather than creators, managers rather than builders."
annievella.com/posts/the-so...
The Software Engineering Identity Crisis - Annie Vella
Many of us became software engineers because we found our identity in building things. Not managing things. Not overseeing things. Building things. With our own hands, our own minds, our own code.
But...
annievella.com
March 26, 2025 at 10:59 PM
I stumbled across this essay as I was falling asleep last night, and left myself multiple emphatic reminders to remember to circle back to it today.
"The joy we're about to lose...we'll become overseers rather than creators, managers rather than builders."
annievella.com/posts/the-so...
"The joy we're about to lose...we'll become overseers rather than creators, managers rather than builders."
annievella.com/posts/the-so...
Reposted by Liz Keogh
We need Wranglers + Engineers.
medium.com/mapai/from-f...
medium.com/mapai/from-f...
From Fair Haven to Technomic Empires.
Why the fuss about conversational programming? Part IV
medium.com
March 26, 2025 at 11:29 PM
We need Wranglers + Engineers.
medium.com/mapai/from-f...
medium.com/mapai/from-f...
Reposted by Liz Keogh
This seems like a good time to remind folks of what Hannah Arendt once wrote:
March 25, 2025 at 1:32 AM
This seems like a good time to remind folks of what Hannah Arendt once wrote:
What gets me about the Signal chat isn't the chat itself.
It's that nobody can answer the question: "What other information has been compromised?"
This will not be the only thread, possibly not the only leak, and anything > 4 weeks old is deleted.
They have no idea what they leaked, to who.
It's that nobody can answer the question: "What other information has been compromised?"
This will not be the only thread, possibly not the only leak, and anything > 4 weeks old is deleted.
They have no idea what they leaked, to who.
March 25, 2025 at 12:31 PM
What gets me about the Signal chat isn't the chat itself.
It's that nobody can answer the question: "What other information has been compromised?"
This will not be the only thread, possibly not the only leak, and anything > 4 weeks old is deleted.
They have no idea what they leaked, to who.
It's that nobody can answer the question: "What other information has been compromised?"
This will not be the only thread, possibly not the only leak, and anything > 4 weeks old is deleted.
They have no idea what they leaked, to who.
@RonJeffries.mastodon.social.ap.brid.gy Just found my copy of Kate Oneal's "The Mythical Italian Restaurant" which you gave me as a gift the first time we met (Agile 2005?) for my clumsy explanation of BDD to you and Chet. Finding it today made me smile - thank you!
March 2, 2025 at 2:53 PM
@RonJeffries.mastodon.social.ap.brid.gy Just found my copy of Kate Oneal's "The Mythical Italian Restaurant" which you gave me as a gift the first time we met (Agile 2005?) for my clumsy explanation of BDD to you and Chet. Finding it today made me smile - thank you!
Don’t ask Amazon Q Developer to put code under test for you. It thinks you’re trying to “put it under”, refuses, and directs you to the Responsible AI Policy instead.
February 6, 2025 at 1:43 PM
Don’t ask Amazon Q Developer to put code under test for you. It thinks you’re trying to “put it under”, refuses, and directs you to the Responsible AI Policy instead.
Reposted by Liz Keogh
Newsletter: Our digital lives are unregulated growth-hacked ecological disasters. The modern tech ecosystem's growth-at-all-costs mindset is actively harming billions of people, pushing them toward authoritarianism for profit. We must fight back.
www.wheresyoured.at/never-forgive-them/
www.wheresyoured.at/never-forgive-them/
Never Forgive Them
In the last year, I’ve spent about 200,000 words on a kind of personal journey where I’ve tried again and again to work out why everything digital feels so broken, and why it seems to keep getting wor...
www.wheresyoured.at
December 16, 2024 at 5:31 PM
Newsletter: Our digital lives are unregulated growth-hacked ecological disasters. The modern tech ecosystem's growth-at-all-costs mindset is actively harming billions of people, pushing them toward authoritarianism for profit. We must fight back.
www.wheresyoured.at/never-forgive-them/
www.wheresyoured.at/never-forgive-them/
Wow, the Cynefin.io wiki has come a long way since I last looked at it! This page on the vector theory of change is really concise and has a great example of dispositional landscapes: cynefin.io/wiki/Vector_... - would love to pass this on to change agents everywhere!
Main Page
The cynefin.io, Naturalising Sense-Making wiki that almost anyone can edit.
Cynefin.io
December 19, 2024 at 12:48 PM
Wow, the Cynefin.io wiki has come a long way since I last looked at it! This page on the vector theory of change is really concise and has a great example of dispositional landscapes: cynefin.io/wiki/Vector_... - would love to pass this on to change agents everywhere!
Here's a puzzle. How do y'all work out what data you need from your users and data constraints, alongside your emergent product, when you can't get the data later if you didn't get it from the get-go and you can't put data constraints in if the data is already a mess? What should I read / watch?
December 11, 2024 at 8:15 AM
Here's a puzzle. How do y'all work out what data you need from your users and data constraints, alongside your emergent product, when you can't get the data later if you didn't get it from the get-go and you can't put data constraints in if the data is already a mess? What should I read / watch?
Had a lot of conversations this year about Doug Hubbard's method of calibrating estimation confidence - found a great write-up of it here. Next time they ask you for an estimate, ask what level of confidence they're looking for!
www.cleverworkarounds.com/2009/11/11/a...
www.cleverworkarounds.com/2009/11/11/a...
A simple way to improve your estimating (and a cool pub trick) – Conclusion
…and we’re back! Well… that was a long commercial break wasn’t it :-) In case you missed part 1 of our version of the show “deal or no deal”, you missed the big cliff-hanger and you really should r…
www.cleverworkarounds.com
December 9, 2024 at 11:35 PM
Had a lot of conversations this year about Doug Hubbard's method of calibrating estimation confidence - found a great write-up of it here. Next time they ask you for an estimate, ask what level of confidence they're looking for!
www.cleverworkarounds.com/2009/11/11/a...
www.cleverworkarounds.com/2009/11/11/a...
Reposted by Liz Keogh
This seemed inevitable to me after just a couple of weeks evaluating GPT-4 and previous versions. First rule of reality: it's *always* an S-curve.
garymarcus.substack.com/p/confirmed-...
garymarcus.substack.com/p/confirmed-...
CONFIRMED: LLMs have indeed reached a point of diminishing returns
Science, sociology, and the likely financial collapse of the Generative AI bubble
garymarcus.substack.com
November 10, 2024 at 6:56 AM
This seemed inevitable to me after just a couple of weeks evaluating GPT-4 and previous versions. First rule of reality: it's *always* an S-curve.
garymarcus.substack.com/p/confirmed-...
garymarcus.substack.com/p/confirmed-...
Reposted by Liz Keogh
@lunivore.bsky.social coined the term 'sadlines' and i think it's revolutionary.
October 24, 2024 at 4:44 PM
@lunivore.bsky.social coined the term 'sadlines' and i think it's revolutionary.
Reposted by Liz Keogh
Can you imagine going to a restaurant and yelling at the waitress to bring your food faster, following her around the room shrieking at her and when she finally sets your in meal in front of you, sniffing it and walking outside. MY CAT CAN
October 17, 2024 at 11:35 AM
Can you imagine going to a restaurant and yelling at the waitress to bring your food faster, following her around the room shrieking at her and when she finally sets your in meal in front of you, sniffing it and walking outside. MY CAT CAN
Reposted by Liz Keogh
12 years later and Cynefin for everyone is still such a useful post. Thanks @lunivore.bsky.social
lizkeogh.com/cynefin-for-...
lizkeogh.com/cynefin-for-...
Cynefin for Everyone!
Five years ago, around Christmas 2012, I wrote an article about Cynefin, the sensemaking framework. I focused it on software development, because that was the main industry I worked in, and particular...
lizkeogh.com
September 16, 2024 at 4:19 PM
12 years later and Cynefin for everyone is still such a useful post. Thanks @lunivore.bsky.social
lizkeogh.com/cynefin-for-...
lizkeogh.com/cynefin-for-...
Reposted by Liz Keogh
Optimist: the cup is 1/2 full
Pessimist: the cup is 1/2 empty
Excel: the cup is January 2nd
Pessimist: the cup is 1/2 empty
Excel: the cup is January 2nd
August 29, 2024 at 12:40 PM
Optimist: the cup is 1/2 full
Pessimist: the cup is 1/2 empty
Excel: the cup is January 2nd
Pessimist: the cup is 1/2 empty
Excel: the cup is January 2nd
Reposted by Liz Keogh
The National Gallery in London is renovating its Sainsbury Wing and they’ve just found a secret letter from one of the original donors, sunk into a concrete column, saying that he hates the columns and is glad they’re being demolished.
10/10 unhinged rich man behaviour, no notes
10/10 unhinged rich man behaviour, no notes
August 27, 2024 at 1:58 PM
The National Gallery in London is renovating its Sainsbury Wing and they’ve just found a secret letter from one of the original donors, sunk into a concrete column, saying that he hates the columns and is glad they’re being demolished.
10/10 unhinged rich man behaviour, no notes
10/10 unhinged rich man behaviour, no notes
Reposted by Liz Keogh
wrote a thing idly after the kids went to bed
August 30, 2024 at 1:33 AM
wrote a thing idly after the kids went to bed
Reposted by Liz Keogh
I'm watching some folks reverse engineer the xz backdoor, sharing some *preliminary* analysis with permission.
The hooked RSA_public_decrypt verifies a signature on the server's host key by a fixed Ed448 key, and then passes a payload to system().
It's RCE, not auth bypass, and gated/unreplayable.
The hooked RSA_public_decrypt verifies a signature on the server's host key by a fixed Ed448 key, and then passes a payload to system().
It's RCE, not auth bypass, and gated/unreplayable.
This might be the best executed supply chain attack we've seen described in the open, and it's a nightmare scenario: malicious, competent, authorized upstream in a widely used library.
Looks like this got caught by chance. Wonder how long it would have taken otherwise.
Looks like this got caught by chance. Wonder how long it would have taken otherwise.
Woah. Backdoor in liblzma targeting ssh servers.
www.openwall.com/lists/oss-se...
It has everything: malicious upstream, masterful obfuscation, detection due to performance degradation, inclusion in OpenSSH via distro patches for systemd support…
Now I’m curious what it does in RSA_public_decrypt
www.openwall.com/lists/oss-se...
It has everything: malicious upstream, masterful obfuscation, detection due to performance degradation, inclusion in OpenSSH via distro patches for systemd support…
Now I’m curious what it does in RSA_public_decrypt
March 30, 2024 at 5:13 PM
I'm watching some folks reverse engineer the xz backdoor, sharing some *preliminary* analysis with permission.
The hooked RSA_public_decrypt verifies a signature on the server's host key by a fixed Ed448 key, and then passes a payload to system().
It's RCE, not auth bypass, and gated/unreplayable.
The hooked RSA_public_decrypt verifies a signature on the server's host key by a fixed Ed448 key, and then passes a payload to system().
It's RCE, not auth bypass, and gated/unreplayable.
Reposted by Liz Keogh
This new data poisoning tool lets artists fight back against generative AI
The tool, called Nightshade, messes up training data in ways that could cause serious damage to image-generating AI models.
https://www.technologyreview.com/2023/10/23/1082189/data-poisoning-artists-fight-generative-ai/
The tool, called Nightshade, messes up training data in ways that could cause serious damage to image-generating AI models.
https://www.technologyreview.com/2023/10/23/1082189/data-poisoning-artists-fight-generative-ai/
This new data poisoning tool lets artists fight back against generative AI
The tool, called Nightshade, messes up training data in ways that could cause serious damage to image-generating AI models.
www.technologyreview.com
October 24, 2023 at 7:37 AM
This new data poisoning tool lets artists fight back against generative AI
The tool, called Nightshade, messes up training data in ways that could cause serious damage to image-generating AI models.
https://www.technologyreview.com/2023/10/23/1082189/data-poisoning-artists-fight-generative-ai/
The tool, called Nightshade, messes up training data in ways that could cause serious damage to image-generating AI models.
https://www.technologyreview.com/2023/10/23/1082189/data-poisoning-artists-fight-generative-ai/