Campfire AI

GPT-4o mini has launched yesterday evening – Open AI's latest GPT model. Here's the breakdown: GPT-4o mini delivers nearly the same level of intelligence as GPT-4o, but in a much smaller, more efficient model. This means: - Cost-Effective: The smaller model size translates to lower operational costs while maintaining performance. - Faster Responses: Latency is significantly reduced due to the smaller model size. - Improved Accuracy: OpenAI claims a 30% reduction in false outputs, which significantly reduces risk. - High Performance: Despite its compact size, GPT-4o mini's performance is comparable to GPT-4o. Overall, this means top-of-class Gen AI just got a lot more accessible and scalable. Combined with improved latency and accuracy, the User Experience of Gen AI applications just got a lot better, while reducing risk. Staying ahead with the latest AI technology is crucial in today’s fast-paced business environment. Want to integrate GPT-4o mini into your business? Contact us to learn how we can help you leverage this cost-efficient model. #GPT4oMini #EnterpriseAI #ConversationalAI

“AI is changing everything, but it's HR that's in for the biggest change.” Alexis at HR Square's HRTech event at Flagey The seminar, coupled with the magazine, focused on the intersection of Tech and HR, emphasizing the importance of keeping the human at the center of high-tech solutions, especially in HR. Since this current wave of AI is very language-heavy, just like HR, the HR field is in for an enormous transformation in the years to come. Leadership will be challenged to adapt fast. With big opportunities come big challenges! Thanks to Margot Wuillaume from eBloom, Stéphanie Reniers 🍀 from Gentis and Frédéric Naedenoen from LENTIC HEC Liège for their insights. And thanks to Fernand Letist for the invitation. #HRTech #TechnologyAndHR #Leadership #FutureOfWork

40% of Belgians (aged 16 to 74) lack basic digital skills. Digital Inclusion is crucial for everyone – citizens, the economy, and government. Together with 63 other companies in Belgium, we've endorsed DigitAll's mission to tackle the digital divide – and call on our governments to act on three recommendations: 1/ Develop a coordinated strategy 2/ Provide structural funding 3/ Work on an inclusive approach We invite you to read and share the open letter to help raise awareness and drive structural change. 🔗 Download the French open letter here: https://lnkd.in/eGeEnePu 🔗 Download the Dutch open letter here: https://lnkd.in/eZkEt4dP #DigitAll #DigitalInclusion #Belgium  Ali Chari Linde Verheyden Catherine Bals Kristof Lambert

What an amazing video! It's not always easy to showcase the work we do, but this story truly shows the tremendous value the voicebot we built together provides. Proud to be a part of this fantastic project! 🎉 We started working with Moet ik naar de dokter? one year ago, and since then, we've built 2 different voice bots in Dutch and English. HAP and DAG (the names of the bots) have been instrumental in speeding up patient triage and reducing waiting times for urgent cases! Special shoutout to 🤖 Geertina, Sophie, Brecht, Sergej and Koen for their hard work on this project. And a big thank you to Vcare connect for being such a great partner! Check out the video below to see the impact of our collaboration.

Campfire has officially partnered with Ebbot! Ebbot's conversational AI platform seamlessly combines traditional intent-based AI with Generative AI. Coupled with their ease of use and highly cuatomizable flows, we believe it's an excellent choice for some of our clients. We have a strict platform-vetting procedure and only work with the very best platforms for our clients 🙏🏼. With a joint project live and more underway, we're looking forward to the value this partnership will bring to our clients. This week we welcomed Gustav, Pontus and Mike from Ebbot to our BeCentral office, to solidify our teamwork. Here's to more Chatbots that actually work! #ConversationalAI #GenerativeAI #CustomerSupportAutomation

Don't rule out old tech! SMS is getting a major update this September when Apple introduces RCS as part of iOS 18. RCS (Rich Communication Services) brings rich media elements like branded profiles, buttons, and carousels to SMS. According to Google, users are 35x more likely to read RCS messages than emails and 74% more likely to engage with brands through RCS! Check the link in the comments to learn more. Want to 35x your read rates with RCS? DM us to chat! (Alexis Safarikas Ivan Westerhof) PS: We're secretly launching the first RCS use case in Belgium next month. Stay tuned for updates 👀 #ConversationalAI #Apple #RCS #ios18 #wwdc2024 📍 Based in Brussels, working globally 🌎

👋 Say hello to Mauricio Userralde! We're thrilled to have Mauricio join us around the Campfire as our newest AI Project Manager! With experience working in The Americas and Europe, he's the perfect human to strengthen our project delivery team. Put your hands up and welcome our newest marshmallow 🙌🔥👏

🔥 Tomas & Alexis will be presenting how we use NLP & LLMs to solve business problems and lead successful Conversational AI projects to the CENTAL seminar tomorrow. Very happy to present Campfire to a group of students and PhDs in the field of A.I. We are so glad to see more and more universities offering master's degrees in NLP, AI, data science, ... 👏 It's so important for us to meet the future talents and show them there are plenty of opportunities for real solutions. Thank you Eva Rolin for the invitation 🙏 #nlp #llm #uclouvain

🔥 Last week, Alexis Safarikas was interviewed by L'Echo regarding the direct impact of the EU's AI Act on Belgian companies. While the impact on Campfire AI is relatively limited, we need to be careful with the use cases we are working on and the way we apply the principles of conversational AI. The main element of this new piece of legislation: defining the level of risk. For Alexis, providing a safe and transparent experience to the end user is key to mitigate this risk. Following UX principles we have been advocating for since the beginning. More information on EU's risk classification 🔎 https://lnkd.in/euaiact Link to the podcast bellow 👇 #euaiact #generativeai #userexperience 📸 Credit: trail for the visual

Google vs. OpenAI vs. Microsoft 🤼♀️ Which is the best?! We’re still in the preliminary testing stages, but we’ve been busy the last ~2 weeks testing Google’s Gemini 1.5. Swipe or scroll to read our first key takeaways and how we think Gemini stacks up against the competition 👇 Google’s Gemini 1.5 ✅ Best pricing $20/month per person with access to 2TB of Drive storage ✅ Largest context-length Ability to digest an entire book, while other models get confused above 10M tokens (1 word = 1.3 tokens in English / 1 word = 2 tokens in Spanish) ❌ Overly censored By trying to eliminate bias, the LLM swings to the other extreme by responding with “incorrect” answers that are the opposite of the stereotypical, yet “correct” answers ❌ Weakest image generation Related to the extreme censorship, it generates inaccurate images OpenAI’s GPT-4 ✅ Smartest general LLM (plug-and-play) Able to solve logic puzzles and questions when other models can’t ✅ Easy to customize GPTs allow you to build no/low-code assistants for specific use cases ❌ Outages 3 outages in Feb. 2023 ❌ Generates unrequested content Creates unrequested images or provides additional content that wasn’t in the original prompt Microsoft 365’s Copilot ✅ Most secure and compliant GDPR-compliant, with servers throughout the world ✅ Enterprise solution Integrated with the entire 365 software offering ❌ Most expensive $30/month per person with an annual commitment ❌ Temperamental  Randomly closes conversation (even when not insulted) Anthropic’s Claude 3 ❌ Limited availability  Right now Claude 3 is only available in certain regions, and Belgium isn’t one of them. We’ll update you with our findings once we can access Claude 3! 📢 As we continue test, we’ll continue to share updates of what we uncover! Stay in the loop by following our page! ⏭ Next Week: Which LLM will pass all the benchmarks?!