@Tasty_Bytes_htx
Fascinating ideas , very excited right now
@JetSoftProHQ
Scaling laws + reasoning = AGI? Fascinating take by Bob McGrew! The challenge is not just computing but making AI agents reliably reason in unpredictable environments. How do we ensure AI remains controllable as it scales? At JetSoftPro, a software development service, we work with AI-driven solutions, and this is THE question we keep running into
@dronebrains9027
Robotics has been limited by control systems for decades. We'll get there soon. But outputting text is useful too
@aimjs
Start with the best models, be at frontier and get it to work 🥊- all else can come after we know where the real value is!
@gustavojavier2353
And when cameras were invented, artists were free from the constraints of depicting reality, and new artistic movements were born such as Cubism, impressionism and so on.
@Hopyboby
Bob is amazing
@brucem8448
I need my Diana Hu content.
@spdwivedi5925
Great Video
@yilei1051
26:42 it is impossible for that to happen. The difference between art and programming is that art is not a high-demand economic value creating job, but programming is. So if programming becomes a form of art (you only program if you think you can write code that others would print and put on walls), then the programming market will shrink to the size of painting market. It won't die but it will decline a lot.
@prao4603
No mention of the elephant in the room (or should I say whale) at all ? - i.e. DeepSeek?
@senju31
Does YC still accept persons if all they have is an idea without a MVP but you have a great understanding/plan of how to implement it?
@CoderThings-d3z
Next batch?
@jrmydnls
I've heard that YC is shifting away from investing in Africa. As a Nigerian founder building a product for the global market, do I still have a chance of getting into YC?
@spacebetweenideas
Imagine you had a basic understanding of what a manager of a team or even a manager of ones own life really needs. What would that be? The challenge with AI is that it's waiting for you to give it a prompt. All these people are creating prompts to solve their problems and give them access to output that they desire. What happens when you distill those behaviours into something like metadata. What are the 'models of thinking' that people need in order to get better output? I would suggest that you start with the objective. What is the objective? It can be more complicated because in a business there are generally 4 layers of objectives. It starts with the vision, then the people/tech capabilities, then the plan, then the performance. Depending on what level of the organization you manage, you're going to define objectives. That's how you begin game one. So you build a UI of an arcade. There are a few foundational games. The first game requires you to express an objective. Naturally, there are three things that follow - work/business services, skills/tools, and people/cooperation. That's is. That's the name of the game. You have the 'parts of and pieces' that you're going to put into the kanban. Who, what, when, status. That's the model for game one. The work orders are all based on the objective. Game two is slightly different because it's a situation where you have an objective at the higher level. Maybe you have a vision but you're still making a decision about capabilities. Or, the case that you mentioned in the video, where you have the AI but you need a UI, a.k.a. a plan. (Game one was a plan.) To get to the plan, you have a UX that takes you through the process of expressing context and defining scenarios. The trick there is that sometimes you trip up comparing scenarios that are incomparable. Like when you need need to do both or all three scenarios. In that case, you take one scenario and it becomes part of the context, until you've drilled down to scenarios where you can compare similarities and differences. There a few more steps, but you get the idea. You're building a pyramid of choices. This vision, these capabilities, these plans, this performance. Flushing out all the decision trees until people are debating the final decisions of branches and leaves. Game one - know what to do and flush it out and assign it and do it. Game two - where you get to the point where there is disagreement about the objective, the work, the skill, or the person... or when. Game three ... Well, I already used Tom Buckholtz work to build game 2. It's one of his Direct Outcomes checklists. Game three uses his State Inventiveness checklist to first consider how others are doing it, particularly those in a leading position, then understanding how to do the same thing or something better likely pulling things apart or pushing them together. Check the connections. There are a couple of more steps but you get the point. Getting people to focus on the categories of thought that actually matter makes the difference in selecting the right objectives to pursue. Logic games that are based on best practices for how high performing humans coach themselves and others to get things done. Behavioural psychology based on a logical foundations of what people do, not feel, when making choices. What's up with AI? It's the same thing I'm doing now. I facilitate these experiences with people all the time. It would be nice if I had someone to talk to that could facilitate the UX for me. Automate. Those are there 3 foundational games that I'm designing. I found a tutor to help me learn to use AI to create code, but first you have to design the UI. At some point, I have to learn to work with others because what's difficult for one is easy for another. Reduce time to market or lose the opportunity. What would you trade for the right person to walk into your life? 😘
@TheBhushanJPawar
I am 41 year old... Am I eligible?
@mykpono
Credits and references are important...
@jijuna4
Ok, they use Dota to help IAs 😂
@Abhijit_journey
i have also a great idea. but unable to create MVP alon need a team a mentor for this idea and much more important is funding. if possible please reply on this comment.....
@yuviiiiiiiiiiiiiiiii
Not useful, nothing new.
@chapterme