rohit89

This depends on the actual number of undiscovered bugs still in curl. If there is nothing to find then even a 10x better Mythos will find nothing. Also I think the quality of the codebase matters a lot when it comes to finding bugs. Its possible that the curl is so well written that it is relatively straightforward for existing ai tools to find bugs.

that hasn't been solved before a bunch of times in publicly available code?

And this matters because? Most devs are not working on novel never before seen problems.

Sam is an investor in a fusion startup. In any case, how long it takes us to get to working fusion is proportional to the amount of funding it recieves. I'm hopeful that increased energy needs will spur more investment into it.

It would not always be a decade or two away if it was funded like the current AI cycle.

Are they training only on video game data though? I would be surprised when its so easy to generate proper training data for this.

It is interesting to think about. This kind of training and model will only capture macro effects. You cannot use this to simulate what happens in a biological cell or tweak a gravity parameter and see how plants grow etc. For a true world model, you'd need to train models that can simulate at microscopic scales as well and then have it all integrated into a bigger model or something.

As an aside, I would love to see something like this for the human body. My belief is that we will only be able to truly solve human health if we have a way of simulating the human body.

So what is final state here for us?

I think we have a long way to go yet. Humanity is still in the early stages of its tech tree with so many unknown and unsolved problems. If ASI does happen and solves literally everything, we will be in a position that is completely alien to what we have right now.

How to even find the value in living given all of that?

I feel like a lot of AI angst comes from people who place their self-worth and value on external validation. There is value in simply existing and doing what you want to do even if nobody else wants it.

The goal is to eventually be able to model physics and all the various interactions accurately.

In the future, everyone will have their own ai agents capable of generating music to their own tastes. They won't be using spotify.

The future with AI is not going to be our current world with some parts replaced by AI. It will be a whole new way of life.

Yea. I feel likes its a waste of time and energy to read/argue about it. Either it works or it doesn't. Everyone already has exposure to it and opinions on it. There's no need to convince anyone. Reality will bear out the results.

If a product has no team assigned there's nobody to do that work.

This seems like a good eval case for autonomous coding agents.

Crypto is a tech that solves problems that most people don't care about.

VR/AR is tech that is nowhere near ready so its premature to make judgements there. I am bullish on XR.

In terms of value-add, many times technology needs to be invented first before we decide whether its worthwhile. Nobody asked for computers in every home, smartphones in every pocket etc.

Short of it being able to mind read, you need to communicate with it in some way. No different from the real world where you'll have a harder time getting things done if you don't know how to effectively communicate. I imagine for a lot of popular use-cases, we'll build a simpler UX for people to click and tap before it gets sent to a model.

1. you're holding it wrong

When computers/internet first came about, there were (and still are!) people who would struggle with basic tasks. Without knowing the specific task you are trying to do, its hard to judge whether its a problem with the model or you.

I would also say that prompting isn't as simple as made out to be. It is a skill in itself and requires you to be a good communicator. In fact, I would say there is a reasonable chance that even if we end up with AGI level models, a good chunk of people will not be able to use it effectively because they can't communicate requirements clearly.

I think there is plenty of demand for software but not enough economic incentive to fulfill every single demand. Even for the software that is being worked on, we are constantly prioritizing between the features we need or want, deciding whether to write our own vs modifying something open source etc etc. You can also look at stuff like electron apps which is a hack to reduce programmer dev time and time to market for cross platform apps. Ideally, you should be writing highly performant native apps for each.

IMO if coding models get good enough to replace devs, we will see an explosion of software before it flattens out.

These are chores and you are vastly underestimating the time saved. The 5-10 min saved per task, they all stack up. Also eventually these would be open source models that you can host yourself so you wouldn't need to worry about giving control to any corporation.

The Orion is the proof of concept. The metaverse that Mark is thinking of is as ambitious as Musk's Mars plans. And it is something that requires large amounts of capital and time.

I do not know where this notion that Meta has given up on the metaverse comes from. Mark continues to talk about AR/VR at every opportunity and Reality Labs continues to invest big on it. The metaverse is a bet but its a 10 year bet that has not been played out yet.

Lisa is smart but this only got to her because George got publicly upset.

If his grievances actually got through to the AMD CEO, I'd say he's already had a bigger impact than most.

Comma and openpilot are doing fine and don't need George anymore. Don't think they will "win self driving cars" though.

Most likely people who predict failure will say of course Mark managed to make it work because he could afford to throw billions on top of billions after it.

The metaverse is a 10 year bet that involves progress across multiple axes. No-one here can predict the state of progress in ten years.

How do you discover new podcasts to index?

Did you use the sentence-transformers model as-is or did you need to fine-tune it for medical data?

In fact, N628TS is airborne over the Middle East right now using its original assigned ICAO hex code of A835AF, not a PIA code.

He's at the football world cup final right now.

If you are really good at something, people will like you inspite of your drawbacks however big they may be.

This take is parroted a lot but makes no sense because he already has a ready made excuse for selling shares with spacex.

Wouldn't it be prohibitively expensive for you to crawl and index the web?

One thing that bothered me a lot in that book was the trio hiding out in the wilderness scrounging for food. They could do magic! How hard was it to sneak into a muggle supermarket and steal some food?! Or lie low at some pre-arranged safe houses? Could have had Dobby bring supplies.

I don't know how effective Taboo ambush tactics would be in the long run but not finishing off death eaters and letting them respawn was definitely a bad move tactically. Should have atleast dumped them all in a fidelius charmed warehouse or something in a Stunned state.

death lurking around the corner every day.

It will still be lurking around the corner and it will change your risk calculus completely. Today, if you go out and get hit by a bus, at worst, you lost 70 years of your life. In the future, you're risking 10-100k years of your life. What's worse is having to accept this risk for your loved ones.

A better question to consider would be, what would people do if they had only 1/5/10 years to live.

From what I see, his system seems to be about focusing on one company, one problem at a time. Neuralink and Boring are side projects. And I believe he definitely doesn't recommend running two companies at a time.