Marc Khoury

@[email protected]
41 Followers
82 Following
603 Posts
AI Software Architect 💻. Degree in Literature and Philosophy 🎓.
Fantasy, sci-fi, digital rights 📚. Talk to me about code, mind, knowledge, learning, life, love, thinking, friends, experience.
Githubhttps://github.com/koganei
Bloghttps://blog.mkhoury.org
LocationQuebec, Canada
Marc KhouryFeb 18, 2024
grantpotterFeb 18, 2024
It seems someone should check first to get a better understanding about what may need to be smashed and what we should avoid smashing. #llm #ai https://electricarchaeology.ca/2024/02/12/smash-the-looms/
Smash the Looms

In which I explain why I’m spending all this time exploring how LLMs might be usefully ensnared for dignified ends. I’m a digital archaeologist. One of the implications of that phrase i…

Electric Archaeology
Marc KhouryFeb 15, 2024
Johan HalseFeb 15, 2024
A thing that just kind of happened as the years went by is that I stopped thinking “what if that was me” and started thinking “what if that’s my boy”
Marc KhouryFeb 13, 2024

We need to build a Software Commons so that we can give everyone the ability to adapt their digital lives to their liking. So that these intimate, private devices to which we entrust most of our attention, these things which have great effects on our cognitive & emotional functions, remain ours in a real sense. We need to create the tools and processes to allow anyone to make modifications to their software by simply expressing that intent.

https://blog.mkhoury.org/post/2024/02/12/free-to-be-weird-lowering-barriers-to-open-source.html

#opensource #llm #ai

Free to be Weird: Lowering barriers to Open Source contributions

Public code repositories like Github are currently being beset by a flood of LLM-generated contributions. It’s becoming a bit of a problem and is one of the facets of the Great Flood the web is currently experiencing.

Digital Dilemma
Marc KhouryFeb 11, 2024
InautiloFeb 11, 2024

#Design #Reports
Apple releases ‘MGIE’ · An AI model for instruction-based image editing https://ilo.im/15xy72

_____
#AI #Chatbot #OpenSource #ImageEditor #ProductDesign #UiDesign #VisualDesign #WebDesign #Content #Image

Marc KhouryFeb 11, 2024
❄️ freezr ❄️Feb 10, 2024

I am using an #AI logo generator to get some inspiration but the design is boring and results are all similar...

My supervisor introduced an AI driven copywright method to add description to the company webstore, it looks handy but the tone won't be dissimilar from another webstore that uses the same back-end.

With the #AI #ML #LLM we introduced another grade of homologization into the #web, after the introduction of the SEO strategies... 

Marc KhouryFeb 10, 2024
Show threadSunnJax 🌷🏳️‍🌈 Feb 10, 2024
@mkhoury I agree, and I keep a closer eye on it than I used to. I also send “thank you’s” to companies and organizations that are doing the right things, letting them know how much I appreciate it.
Marc KhouryFeb 9, 2024
Laurie VossFeb 9, 2024
A fun little edge-case test for multi-modal LLMs: ask it to generate "four interlocking puzzle pieces". I tried with both Gemini Ultra and GPT-4. It took several revisions in each case; Gemini eventually did a fine job but GPT just can't seem to figure it out.
Marc KhouryFeb 8, 2024
The FrogFeb 8, 2024

It's official, "enshittification" officially now has a french equivalent: "merdification"

@pluralistic
#enshittification
#merdification

https://ici.radio-canada.ca/nouvelle/2047650/merdification-enshittification-cory-doctorow-decrypteurs

La « merdification » du web, avec Cory Doctorow

De plus en plus d'internautes trouvent que les services web dépérissent, ce que le blogueur canadien Cory Doctorow qualifie de « merdification »

Radio-Canada
Marc KhouryFeb 6, 2024
Women Interface Network  Feb 6, 2024
👾 I really contest the much stated notion that fedi is so unsafe for minorities, and especially that twitter or bluesky are better places

here we can actually have control over our spaces, we can take matters into our own hands, we aren't dependent on some underpaid overworked faceless moderation team or worse, an automated system.

it's always managerial class tech people promoting such ideas. it's the same old liberal argument against anarchism, it's just the fear of freedom, the fear that we couldn't possibly keep ourselves safe and need corporations to do it for us.
Marc KhouryFeb 2, 2024
Show threadMatti SchneiderFeb 2, 2024
(and big up too to #OLMo 7b, an #LLM published yesterday with all its pre-training data, training and evaluating code, and trained on #EuroHPC 🇫🇮 compute. “Open weights” is not #OpenSource. https://blog.allenai.org/hello-olmo-a-truly-open-llm-43f7e7359222 )
#EUOpenSource
Hello OLMo: A truly open LLM - AI2 Blog

As the world races to deploy AI models that are effective and safe, the demand for Open Large Language Models (LLMs) has exploded. The massive adoption of both open and closed AI models means that AI…

AI2 Blog