By "YouSum Live" 00:05:12 Clara's AI assists in dispute resolution 00:06:32 Clara's AI implementation improves decision-making 00:08:05 Customer satisfaction equals human agent performance 00:11:02 AI enhances customer support efficiency 00:12:39 AI reduces customer service resolution time 00:27:12 AI tools can enhance marketing campaigns 00:30:01 AI-generated content lacks true creativity 00:32:20 Kiki centralizes internal knowledge for employees 00:43:11 Future AI will automate financial advice 00:47:11 AI can personalize shopping experiences By "YouSum Live"
🎯 Key points for quick navigation: 00:00 *🛠️ Initial Development of Klarna's AI* - Introduction to Sebastian and the business context of Klarna, - Klarna's expansion and initial AI implementation. 02:59 *🌐 Early AI Integration Strategies* - Sebastian's initial exposure to AI in 2022, - Interaction with Sam Altman and OpenAI to develop AI use cases at Klarna, - Internal trials and establishment of processes for AI experimentation. 05:04 *⚖️ AI in Dispute Resolution* - Klarna's experimentation with AI in dispute resolution, - Development of co-pilots to assist in complex customer service tasks, - Positive outcomes and initial production implementation. 07:23 *🤖 AI Implementation and Customer Perception* - Transition from co-pilots to client-facing AI, - The importance of transparency in AI-human interaction, - Customer satisfaction and operational efficiency improvements. 10:21 *📖 Detailed Instructions for AI Success* - Challenges of AI documentation and clarity, - Ensuring high-quality data input for effective AI outcomes, - Parallel improvements in human resources alongside AI. 11:56 *📊 AI Performance Metrics* - Reduced response times and improved customer service efficiency, - The impact of AI on job roles and company profitability, - Customer preferences and the evolving role of AI in customer support. 15:38 *📈 Broader Impacts of AI on Workforce* - Potential reduction in workforce requirements due to AI, - Financial improvements for Klarna through AI adoption, - The need for proactive societal measures to manage workforce transitions. 17:54 *🌍 Societal Responsibilities and Future Planning* - Importance of electronic identification to prevent fraud, - Balancing progress with societal impact and job displacement, - Advocating for supportive measures for those affected by AI-driven changes. 21:13 *🤔 Impact of AI on Jobs* - AI's impact on job categories and the societal reactions. - Advice to target unpopular jobs with AI. - Emotional resonance related to job displacement fears. 24:03 *🎨 AI in Marketing* - Creation and adaptation of marketing materials using AI. - Balancing AI and human creativity in campaign generation. - Efficiency and quality improvements in marketing. 28:01 *🧠 AI's Creative Limitations* - Differences in AI's capability in image vs. copy generation. - AI's tendency to produce average results. - Human creativity still surpasses AI in unique content production. 30:15 *📚 Internal Collaboration and Information Systems* - Integration of internal knowledge systems with AI. - Kiki's role in centralizing and accessing company information. - Efficiency and productivity improvements from information standardization. 35:05 *💡 Buy vs. Build AI Solutions* - Decision-making between developing or purchasing AI solutions. - Importance of learning and adapting new technology internally. - Implications of open-source and public APIs on AI integration. 40:09 *🚀 Future AI Innovations and Products* - Upcoming AI-driven customer service enhancements. - Long-term vision of AI-driven financial advice and decision-making. - Potential to disrupt banking industry through AI-enabled customer mobility. 42:51 *🌐 Future Vision of Fintech* - Klarna's vision as a digital financial advisor, - AI's role in streamlining the user's financial management. 44:04 *🛍️ AI in E-Commerce Curation* - Role of AI in product recommendation and selection, - Challenges and potential of integrating AI in commerce. 46:21 *🛠️ Personalized Shopping Experience* - Personalized shopping using AI-generated images, - Example of creating custom product images that appeal to customers. 48:57 *🏆 AI Impact in Various Fields* - Admiration for AI leaders like Sam Altman, - Discussion on AI's role in art and the importance of emotional connection. 50:06 *🚀 Advice for AI Founders & Companies* - Key advice for companies embracing AI, - Encouragement to explore and not fear AI advancements. Made with HARPA AI
Check out Kenneth Stanley work on open ended AI, his seminal book "Why Greatness Cannot Be Planned: The Myth of the Objective" is a good place to start.
My view is the concepts of Agents, Tools etc will be part of a enhanced RPA process and can be incorporated as another capability within a RPA tool. I dont see why this has to be a standalone concept and needs a seperate set of tools and techniques. Definitely LLM is transformative but not concepts like Agents and Tools.
The audience can make up its mind about the mind state of people that say "safety is reliability" and "AGI will do boring tasks for you on your computer".
I am pretty sure that we are going to need agents to curate enterprise data because it is local but there is nothing about coding, testing etc wouldn't be solved at the model level ? Matan himself alluded to this, models getting better gives them better results.
"We had an important part to play in the internet revolutions" uhmmm, but breaking all the competition laws and illegally destroying companies that were making good products and forcing us to use your *(^#. How can you say "We had an important part to play in the internet revolutions" with a straight face?
His comment on where we are in the late 1800s is on point. That was during the Second Industrial Revolution. We're in the beginning stages of the Third Industrial Revolution, which will happen a hell of a lot faster and with much larger effects than even that one did.
Because it's not, type 2 thinking is real, type 1 thinking is what you're talking about, you can build a 2 from a 1 with some loops and more architecture. 1 comes first, 2 is an upgrade. 2 requires something 1 doesn't do, unbounded compute, aka the ability to think longer, because you're planning internally, internal thinking, and comparing your plans to choose the better one, and using logic and induction and deduction to work through them. This is what humans do, well documented, not an illusion, just a different mode.
@@Gnaritas42 perhaps declaring axioms as truths and reverse engineering data to present in a book filled with confirmation bias is not exclusively Kahneman's achievement. This appears to be your type 1 thinking it's type 2 (both are are illusory )
And people who don't know enough about AI wonder why the next models are taking several years and not *one* year. It's because they're working on next-level stuff, people! It takes time!
What is so hard about this obviously we need to leverage the gigantic compute available today and genetic algorithms and train models with loss functions that combine at each step the score in non memorization and memorization tasks together while searching in the model architecture space with the genetic algo, evolution is dumb as a rock and creates human Brains so clearly it’s a brute force problem
By "YouSum Live" 00:00:00 Need for depth in AI research 00:00:15 Importance of solving depth problem 00:01:30 Personal background shaping interest in technology 00:02:54 Inspiration from parents' pursuit of excellence 00:05:00 Transition to interest in physics and root problems 00:06:09 Significance of working on root node problems 00:12:01 Combining learning and search for superhuman agents 00:14:05 Challenges in achieving general superhuman agents 00:19:19 Defining agents in the context of AI evolution 00:40:03 Post-training reinforcement learning enhances model behavior 00:40:19 Reinforcement learning refines neural network actions 00:41:01 Reward models influence reinforcement learning effectiveness 00:51:20 Safety and reliability crucial for AI development 00:57:43 Science of AI akin to early physics discoveries 00:59:22 Operational efficiency and focus key in research 01:01:34 Dedication and humility in pioneering AI 01:01:56 Inspiring leadership driving team dedication 01:02:27 Creating a culture of excellence and innovation 01:03:00 Advice on pursuing meaningful work independently 01:04:27 Importance of internal drive in problem-solving 01:04:51 Customer-centric focus crucial for success 01:05:15 Personal alignment with problem-solving essential 01:05:56 Passion and personal connection vital in entrepreneurship 01:06:00 Acknowledgment of tackling significant challenges in AI By "YouSum Live"
It’s not simple enough to answer in a comment bro 😭 they have a bunch of metrics that are basically useless because the llms keep surpassing them but then there are human evaluation metrics as well, for example, how much sense it makes to a human.
I actually think this is an interesting area of opportunity. I think SWE-Bench is the best benchmark for testing real-world agentic capabilities for programming, amongst those that exist today. But we need more and better benchmarks!
This is the most honest explanation for whole AI situation at the moment: better frontier models are coming. They might not be AGI, but still can have a huge impact on the industry - cheaper, better reasoning, less hallucination and so on. Anyone who architects the product assuming the frontier model will be stuck at today's capabilities, won't be in a good situation.
People went from being blown away at GPT-4 to assuming AI will be stuck at it in the space of literally one year. Even Claude 3.5, which is better, didn't do much to change that perception, because it wasn't astoundingly better. It is the strangest and least justified assumption to make about a rapidly advancing technology in living memory. It's how we are wired as humans, that we are almost impossible to impress, but even so, rationally, it is nonsensical. I don't care if it takes one, two or even three more years for the next major leap in frontier models - if GPT-5 is to GPT-4 what GPT-4 was to GPT-3, it will be a staggering improvement, and we will be entering territory that if it isn't AGI, will be so capable as to render the term meaningless. AGI is such a vague term with so many definitions now that the term is already functionally meaningless - yeah yeah, AGI or not, what can this model DO? AlphaFold 2 has been cited over 20,000 times in papers since it became open-source exactly three years ago, turning medicine on its head. But human trials by their very nature take time. And now AlphaFold 3, which is substantially simpler and better, is out, and who knows what it will do, along with all the other AI models being integrated into medicine right now. But because we don't already have the cure for cancer or whatever, and these are medical models and not sexy chatbots, nobody notices outside of those of us who take a particular interest in such things.
This tells me GPT-4 was the top of the last sigmoid curve, and it will take about a year from now to scale the next generation of models for release. And no, he is not some hack, he DOES know what he is talking about. Check out his Wiki bio.
A year is also very fast if you take into account the change in capabilities of gpt 3 vs gpt 4, gpt 5 would be crazy. And If that be the case I will start believing in this theory of scaling laws
@@nuclearenergy919 Indeed. If it be the case. But really, twice as long as that would still be insanely fast too, again considering the jump between 3 and 4. Even if scaling stopped dead tomorrow, what we already have is enough to massively change society as it is integrated into all kinds of systems over the next decade, and we know that the models are just going to keep getting better in that timeframe anyway, whether it be due to scaling or other methods. There's lots of gas in the tank, scaling is just one tool. Interview was grounded and hype-free. Dude is based af, as others here have said.
Healthcare was my first thought as well. "But it hallucinates!" Yeah, and how many doctors make huge f*ck-ups that destroy patients' lives or kill them every day? Hell, if ChatGPT had been available in 2007, I could have caught and gotten off the benzo addiction that destroyed my life by myself very early on, before it could do any damage. It would have all been over in 4 or 5 months. If not that, my useless piece of sh*t doctor could have used it to figure out what to do a few months later when I showed up at his office heavily addicted, and it would have still all been over in 7 or 8 months. Instead of 17 years. And counting.
co piolt is too hyper filtered. microsoft is a corporation that does not care about out privacy but hes a really neat guy. Lets not forget with their openai partnership and all the sketch behind openai since day one..... then the nsa on the board thats scary. still hes a neat guy.
