The fastest way to eliminate fear is to practice. I had the team go through it one day. They didn't get a choice. I locked us on a screen share until everyone was comfortable with how rebasing works. The call lasted maybe 90 minutes. You just have to decide one day that you (or the team) will master this shit, spend a few hours doing it, and move on.
Rebase is a super power but there are a few ground rules to follow that can make it go a lot better. Doing things across many smaller commits can make rebase less painful downstream. One of the most important things is to learn that sometimes a rebase is not really feasible. This isn't a sign that your tools are lacking. This is a sign that you've perhaps deviated so far that you need to reevaluate your organization of labor.
PPMd is the most exotic compressor I've actually used in production. The first time I saw it in action I thought it was lossy or something was broken. I had never seen structured text compress that well.
In theory the risk is immense and incalculable, but in practice I've never found any real danger. I've run wide open powershell with an OAI agent and just walked away for a few hours. It's a bit of a rush at first but then you realize it's never going to do anything crazy.
The base model itself is biased away from actions that would lead to large scale destruction. Compound over time and you probably never get anywhere too scary.
> LLMs seem to have a poor understanding of 3D space.
This is definitely my experience as well. However, in this situation it seems we are mostly working in "local" space, not "world" space wherein there are a lot of objects transformed relative to one another. There is also the massive benefit of having a fundamentally parametric representation of geometry.
I've been developing something similar around Unity, but I am not making competence in spatial domains a mandatory element. I am more interested in the LLM's ability to query scene objects, manage components, and fully own the scripting concerns behind everything.
If 44.1kHz is otherwise sufficient but you have a downstream workflow that is incompatible, there are arguments for doing this. It can be done with no loss in quality.
From an information theory perspective, this is like putting a smaller pipe right through the middle of a bigger one. The channel capacity is the only variable that is changing and we are increasing it.
That's not how audio works. PCM data at some sample rate with infinite precision samples perfectly encodes all the audio data of all frequencies up to half the sample rate. Resampling is a theoretically lossless operation when the frequency content of the audio fits within half the sample rate of both the source and the destination sample rates (which will always be true when resampling to a higher sample rate, FWIW).
The issues are that 1) resampling has a performance and latency cost, 2) better resampling has a higher performance and latency cost
> Take this seriously. If your iPhone does not have Apple’s new update, you must install it now. We know attacks on iPhones have started. We have been warned the threat will extend well beyond those highly targeted initial attacks. And hundreds of millions of iPhone users are also now facing down an unwelcome surprise.
Building some prototypes around the Recursive Language Model paper. I'm currently working on integrating it with Unity for scene and script automation. Thinking about standardized patterns for retrofitting this into other existing business systems. If I can make it drive Unity reasonably well, I think it could drive a lot of things.
> Test these new tools, with care, with weeks of work, not in a five minutes test where you can just reinforce your own beliefs. Find a way to multiply yourself, and if it does not work for you, try again every few months.
I've been taking a proper whack at the tree every 6 months or so. This time it seems like it might actually fall over. Every prior attempt I could barely justify spending $10-20 in API credits before it was obvious I was wasting my time. I spent $80 on tokens last night and I'm still not convinced it won't work.
Whether or not AI is morally acceptable is a debate I wish I had the luxury of engaging in. I don't think rejecting it would allow me to serve any good other than in my own mind. It's really easy to have certain views when you can afford to. Most of us don't have the privilege of rejecting the potential that this technology affords. We can complain about it but it won't change what our employers decide to do.
Walk the game theory for 5 minutes. This is a game of musical chairs. We really wish it isn't. But it is. And we need to consider the implications of that. It might be better to join the "bad guys" if you actually want to help those around you. Perhaps even become the worst bad guy and beat the rest of them to a functional Death Star. Being unemployed is not a great position to be in if you wish to assist your allies. Big picture, you could fight AI downstream by capitalizing on it near term. No one is keeping score. You might be in your own head, but you are allowed to change that whenever you want.
Wouldn't a lot of us become unemployed anyway if there are 75% less jobs? I don't see how I can use AI better than other people. People who keep their jobs are also not in for a fun time when they will be responsible for 4x the surface. And if you are not in top 7 companies, your company might not fire you but get bankrupt in a couple of years because all the investment is hogged by the top7. This is more of a lose-lose situation.
Fiber runs cool because it's operating well within the physical capacity of the channel. Copper needs an incredibly high signal to noise ratio to overcome the limitations of its medium. Copper will consume 5-10x more power than fiber for the same # of bits transmitted.
Rebase is a super power but there are a few ground rules to follow that can make it go a lot better. Doing things across many smaller commits can make rebase less painful downstream. One of the most important things is to learn that sometimes a rebase is not really feasible. This isn't a sign that your tools are lacking. This is a sign that you've perhaps deviated so far that you need to reevaluate your organization of labor.
reply