AI for Enterprise Orchestration to decide better and move faster.
Brussels, Belgium
Joined October 2017
- Tweets664
- Following195
- Followers756
- Likes180
Superlinear retweeted
Announcing 🥤 RAGLite: a Python package for building Retrieval-Augmented Generation (RAG) applications. RAG can be magical, but is only as good as its retrieval quality. RAGLite's goal is to unhobble RAG by providing the best solutions to each of its subproblems. Link below 👇
Superlinear retweeted
In an interview with @MarietjeSchaake, @mustafasuleyman said that @inflectionAI would be announcing a new LLM in the coming weeks that is trained with 10²⁵ flop. He also predicted that we could reach AGI in 18 months by scaling to 10²⁷ – 10²⁸ flop. Some observations: 👇 (1/4)
For those who missed my conversation with @mustafasuleyman at @StanfordHAI, we talked about generative AI governance, what to expect in the near future, human rights and AI, and what motivated a social worker to become a tech CEO ↘️ m.youtube.com/watch?v=VUOPCI…
Superlinear retweeted
#techeu23 companies don’t need to think about how to use #AI in their workflow.
➡️ They need to RETHINK their workflow #transformation @daviolarnout from @radix_ai
Superlinear retweeted
Hippo was het grootste mysterie uit dit seizoen van de #MaskedSinger
Maar wat dacht onze A.I. agent #ONTMASKATRON?
😁😎
@radix_ai
Superlinear retweeted
Onze stemvergelijkende Masked Singer A.I. #Ontmaskatron zat er nog eens boenk op vorig vrijdag!
Jammer eigenlijk, Sam Gooris was wel cool geweest 😁
@radix_ai
Superlinear retweeted
Hoe deed ons AI systeem ONTMASKATRON het deze week in The Masked Singer?
Well, close… 😁
@radix_ai
Superlinear retweeted
I'm building a company and I'm a father of 2.
I believe that entrepreneurship & technology are the best ways to build a better, safer, more sustainable and healthier future for my kids & humanity.
It is incredibly frustating being in a system that loves status quo so much.
1/7
Superlinear retweeted
ONTMASKATRON is gebouwd door de AI wizards van @radix_ai
Superlinear retweeted
Gisteren "De Afspraak" met gezien met @bartschols en @lucsteels. Te veel om op te sommen waarom ik niet akkoord ben, maar hierbij de samenvatting. 1/5
Superlinear retweeted
Nieuw projectje: proberen de Masked Singer te herkennen met Artificial Intelligence 😁🎤🤓
Kan artificiële intelligentie meespeuren naar #TheMaskedSinger op @VTM ?👀
Volgens @lievenscheire en @radix_ai zeker! 😁Hun "ONTMASKATRON" scoorde al 4 op een rij😶Wij speurden mee met de cel wetenschap & technologie🔍@HLN_BE
hln.be/wetenschap-en-planeet…
#themaskedsingervtm #vtm
Superlinear retweeted
Chinchilla's law predicts that if we can collect twice the amount of training data (which would likely exhaust all human text), the best possible Transformer-based LLM would only require 128B parameters (25% fewer than GPT-3) and could be trained for just $7M. (1/4)
Superlinear retweeted
OpenAI is now making it clear that they want to work towards AGI. They no longer beat around the bush.
Meta open sources the best language model at the moment.
The race has really started now.
openai.com/blog/planning-for…
research.facebook.com/public…
Superlinear retweeted
Companies that don’t have an AI strategy today, brace yourselves!
Incumbents have the advantage when it comes to AI (data and distribution). But startups will redesign workflows from first principles to maximize the value of AI, often creating vastly superior user experiences. It’s very hard for an incumbent to “burn the boats” and throw away its existing UX. Only the very best CEOs will pivot to capture the opportunity.
Superlinear retweeted
Boeiend! Laurent Sorber dient Luc Steels van antwoord (daar is Twitter toch goed voor!). Ik had hun standpunten over generatieve AI heel bewust tegenover elkaar geplaatst en heb uitdrukkelijk zelf geen standpunt ingenomen - deze twee mensen weten er 100 keer meer vanaf dan ik.
Afgelopen weekend schreef @ddeckmyn in een interview met @lucsteels in @DeStandaard dat Luc en ik "een diepe kloof vertegenwoordigen tussen AI-wetenschappers". Wat waren die twee standpunten, en waarom is er onenigheid over? (1/17)
Superlinear retweeted
A new year, a new post!🥳
In our first blogpost of 2023, alumnus @JorenYuuji discusses the interaction & tension points between the #GDPR and #AI from the perspective of an AI practitioner.
Read it here: law.kuleuven.be/ai-summer-sc…
Wishing you a happy new year & happy reading!
Superlinear retweeted
GPT-3 still sounds like a rambling drunk, unable to make a coherent point because it doesn't really understand the world.
But this will change soon says @LaurentSorber
@Siftedeu #AI #GPT3 #MachineLearning
sifted.eu/articles/gpt-3-ai-…
Superlinear retweeted
I am trying to have a view on AI service companies in Belgium:
@BoltzmannML (Ghent)
@BrainjarAI (Leuven)
@ML2Grow (Ghent)
@ml6team (Ghent)
@radix_ai (Brussels/Ghent)
@robovision (Ghent)
The Grain (Antwerp)
There must be a lot more, so kindly add in the comments :-)
Superlinear retweeted
.Well done Davio en team @radix_ai! Keep up the good work. Vlaamse start-up versnelt vaccinontwikkeling bij farmareus GSK | De Tijd tijd.be/ondernemen/farma-bio…
Superlinear retweeted
Is #HR klaar voor AI? Daar gaan onze sprekers @daviolarnout van @radix_ai en @stephenlernout van @nalantis_ai vanavond dieper op in tijdens ons 3e seminarie van #aiacademy. Meer dan 70 AI-believers vonden alvast hun weg naar het Skyline Park in Izegem! #daaromVoka #SamenGroeien
Superlinear retweeted
"The EU has the opportunity to take pole position in the global #AI race if the right policy & regulatory framework is put in place" was the key message shared at Orgalim's roundtable discussion on #IndustrialAI this week. Read about the debate highlights: bit.ly/2LgKYV9
