Unpopular opinion: sorry but it's not the researchers/university professors who build the AI systems used daily by people. It's the engineers. And I think they are the best placed to talk about the use of AI then.
@dfintelligence
-
The dangers of AI attract attention, not its benefits
By
–
Yes, I too first had questions about using AI in my work for certain cases. But if you talk about the dangers of AI, everyone will hand you the microphone. If you talk about the benefits of AI, nobody cares.
-
OpenAI’s hypocrisy: it’s all about money
By
–
A year ago, OpenAI would never have signed something like this. They were the first to push to release hyper models to the public. So this hypocrisy makes me laugh. It's all about money, let's stop lying to ourselves.
-
Tiresome pedagogy against AI misconceptions
By
–
And yes. Every day I have to re-educate clients who hear all sorts of nonsense about AI. That's why it annoys me so much those who just talk for clicks.
-
Sensationalism in AI complicates explaining and applying tech
By
–
Frankly, this kind of sensationalist nonsense in #AI makes our work even harder when we try to explain and apply the technology. It's exhausting.
-
Brute Force Method for Code Generation
By
–
Blog with plenty of examples: https://codi-gen.github.io Paper: https://arxiv.org/abs/2305.11846 Still, be aware that the method is quite brute force and not highly scalable, as seen in the 'test' generations—it often revolves around rain, streets, etc. But it's encouraging.
-
CoDi: A Revolutionary Multimodal AI
By
–
🚨 wow. Générer n'importe quoi à partir de n'importe quoi ?
— Defend Intelligence (Anis Ayari) (@DFintelligence) 23 mai 2023
C'est ce que fait 'CoDi'.
Un des papiers les plus impressionnant que jai vu ces derniers temps.
C'est clairement vers quoi va l'IA aujourd'hui. Des models multimodaux en input et output.
Les liens ⬇️ pic.twitter.com/lMs3g0i367Wow. Generate anything from anything? That's exactly what 'CoDi' does. One of the most impressive papers I've seen lately. This is clearly where AI is heading today—multimodal models for both input and output.
-
Developer workflow shift: reliance on ChatGPT for coding
By
–
Ok. Je peux plus coder sans ChatGPT. C'est réelle. (et c'est trop bien. ) Ca doit faire des semaines j'ai pas ouvert StackOverflow…
-
Public engagement with AI research and latent space concepts
By
–
800 j’aime en mettant le mot « espace latent » et en partagent un papier de recherche 🥲. (Avec une vidéo stylé)
— Defend Intelligence (Anis Ayari) (@DFintelligence) 19 mai 2023
Ça donne de l’espoir. La preuve que la science intéresse peut être… sans besoin d’aller chercher du putaclick angoissant pour parler d’IA. (A bon entendeur) https://t.co/iAPZ6YdyDo800 j’aime en mettant le mot « espace latent » et en partagent un papier de recherche . (Avec une vidéo stylé) Ça donne de l’espoir. La preuve que la science intéresse peut être… sans besoin d’aller chercher du putaclick angoissant pour parler d’IA. (A bon entendeur)
-
Drag your GAN: Interactive Latent Space Manipulation for Image Editing
By
–
Une nouvelle méthode « Drag your GAN » permet d’éditer les photos directement issu de modèles génératif en manipulant directement l’espace latent. Intéressant ☺️.
— Defend Intelligence (Anis Ayari) (@DFintelligence) 19 mai 2023
Le papier : https://t.co/o8Mt7BCAJs pic.twitter.com/42CHVHj0izUne nouvelle méthode « Drag your GAN » permet d’éditer les photos directement issu de modèles génératif en manipulant directement l’espace latent. Intéressant . Le papier : https://
arxiv.org/abs/2305.10973