tippytippytango

Above I was talking more generally about full autonomy. I agree the combined human + fsd system can be at least as safe as a human driver, perhaps more, if you have a good driver. As a frequent user of FSD, it's unreliability can be a feature, it constantly reminds me it can't be fully trusted, so I shadow drive and pay full attention. So it's like having a second pair of eyes on the road.

I worry that when it gets to 10,000 mile per incident reliability that it's going to be hard to remind myself I need to pay attention. At which point it becomes a de facto unsupervised system and its reliability falls to that of the autonomous system, rather than the reliability of human + autonomy, an enormous gap.

Of course, I could be wrong. Which is why we need some trusted third party validation of these ideas.

Ultimately, anecdotes and testimonials of a product like this are irrelevant. But the public discourse hasn't caught up with it. People talk about it like it's a new game console or app, giving their positive or negative testimonials, as if this is the correct way to validate the product.

Only rigorous, continual, third party validation that the system is effective and safe would be relevant. It should be evaluated more like a medical treatment.

This gets especially relevant when it gets into an intermediate regime where it can go 10,000 miles without a catastrophic incident. At that level of reliability you can find lots of people who claim "it's driven me around for 2 years without any problem, what are you complaining about?"

10,000 mile per incident fault rate is actually catastrophic. That means the average driver has a serious, life threatening incident every year at an average driving rate. That would be a public safety crisis.

We run into the problem again in the 100,000 mile per incident range. This is still not safe. Yet, that's reliable enough where you can find many people who can potentially get lucky and live their whole life and not see the system cause a catastrophic incident. Yet, it's still 2-5x worse than the average driver.

It's difficult to do because of how well matched they are to the hardware we have. They were partially designed to solve the mismatch between RNNs and GPUs, and they are way too good at it. If you come up with something truly new, it's quite likely you have to influence hardware makers to help scale your idea. That makes any new idea fundamentally coupled to hardware, and that's the lesson we should be taking from this. Work on the idea as a simultaneous synthesis of hardware and software. But, it also means that fundamental change is measured in decade scales.

I get the impulse to do something new, to be radically different and stand out, especially when everyone is obsessing over it, but we are going to be stuck with transformers for a while.

We need a silver tongued LLM agent that can align all these forces (and a well provisioned MCP paypal tool for greasing palms)

He’s playing the game. You have to say AGI is your goal to get attention. It’s just like the YouTube thumbnail game. You can hate it, but you still have to play if you want people to pay attention.

Sometimes we get confused by the difference between technological and scientific progress. When science makes progress it unlocks new S-curves that progress at an incredible pace until you get into the diminishing returns region. People complain of slowing progress but it was always slow, you just didn’t notice that nothing new was happening during the exponential take off of the S-curve, just furious optimization.

False positives for this tech is absurdly high and law enforcement treats it like it’s perfect. That’s enough of a reason to make it illegal.

There will have to be a new IP framework to make sure people are properly incentivized to produce content in an AI first world.

I hope they can figure out why these give some people headaches and eye strain (like myself) I really want to use this, but can't stand the pain for more than a few minutes.

Sounds like you found a good problem for the students. Having the experience of failing to get the right answer out of the tool and then succeeding on your whits creates an opportunity to learn these tools benefit from disciplined usage.

Patronizing much?

This article captures a lot of the problem. It’s often frustrating how it tries to work around really simple issues with complex workarounds that don’t work at all. I tell it the secret simple thing it’s missing and it gets it. It always makes me think, god help the vibe coders that can’t read code. I actually feel bad for them.

The pedantry in the comments of a todo app is exquisite. HN never disappoints.

Very nostalgic OP, warms my heart 10/10

Sort of, but in a good way, if I’ve spent $15 on a problem and it’s not solved, it reminds me to stop wasting tokens and think of a better strategy. On net it makes me use less tokens, but more for efficiency. I mostly love that I don’t need to periodically do math on a subscription to see if I’m getting a good deal this month.

I prefer just paying for metered use on every request. I hope monthly fees don’t carry over from the last era of tech. It’s fine to charge consumers $10 per month. But once it’s over $50 let’s not pretend you are hoping I under utilize the service, and you want me to think I’m over utilizing it. These premium subscriptions are too much for me to pretend that math doesn’t exist.

I wouldn’t beat yourself up over it. Very few papers can be understood without reading a significant amount of the neighboring literature and the history of how that work came to be. There are norms and customs and a kind of academic language in every community that you won’t be able to see unless you’ve read a lot from that community. Even if you have the right math level it’s tricky.

A single paper is part of a conversation, not something that stands alone. Trying to read one random paper is like finding a 1000 page thread on an obscure topic that has been running for 10+ years and reading only the last page. It won’t make any sense without reading back a ways.

If you pretend/imagine it was intentional, and insightful, you've created a nerd trap for amateur ontologists. Some of which decide to become professional ontologists and sell books on objected oriented design.

The python zoom in seems performative. A vision model already has access to all the data, how does zooming in help it? Still very cool that it can!

These glasses give me an instant headache and 1080p is abysmal if you are used to 5K displays. I love the idea, hate the actual glasses.

It’s fascinating how the debate is going exactly as the car debate went. People were arguing for a whole spectrum of environment modifications for self driving cars.

I’ll take the other side of that bet. The software industry won’t make things easier for LLMs. A few will try, but will get burned by the tech changing too fast to target. Seeing this, people will by and large stay focused on designing their ecosystems for humans.

Whether vibe coding works for you depends on how tight your requirements are.

People expect the llm version of siri to be an agent, not a chat bot. There's a very wide gulf between a chatbot like chatgpt and an agent like... well I haven't found any good agents I'd give write access to my calendar yet so I don't know what to cite.

I had the same experience, it was very fun, but makes it extremely difficult to get excited for corporate career goals after doing something like that.

He needs to learn grit and how to ask for help. He needs to learn some things are hard and that he can’t always lean on his intelligence.

The best way to guarantee a gifted kid wastes a lot of their potential is to be in an environment that is too easy. It creates a devastating mental habit that won’t trigger until later in life, like college. Whenever they try to do something that doesn’t come easy, their brain will try to shut down out of a kind of frustration. They won’t know how to overpower it. It will cause depression, anxiety, shame and low self worth later on. Because the gifted kid will know they are wasting their potential, but blame themself for not being good enough to deal with it. It feels like being broken.

All of this is created by being rewarded for maxing out the rewards of a trivial environment. Someone needs to patiently and compassionately teach them to value overcoming appropriately sized challenges. To find and operate on the edge of their potential and ask for help to operate beyond those limits.

So yeah, grit and asking for help. Intelligence is mostly wasted without it.

Yep, we all have to hit the wall and that’s where we find out what we’re made of. It can be a valuable experience with the right people around to help.

Yep. LLMs can get all the unit tests to pass. But not the acceptance tests. The discouraging thing is you might have all green checks on the unit tests, but you can’t get the acceptance tests to pass without starting over.

A big issue with all this housing wealth is that it's fake. If a good business goes up 10x in value, it accomplishes that by providing more valuable goods and services to people. If a house goes up 10x in value, that could only be achieved by ensuring supply grows slower than demand. The house didn't produce any net positive for society, actually the opposite. The owner is being rewarded for figuring out how to increase demand for their house while providing nothing in exchange to the broader world. It's a serious bug in capitalism that we call this "building wealth" it should be called "building scarcity". Might as well hoard bitcoins.

I don’t think I could say anything that hasn’t already been written about the causes of this problem. But, I’m glad we have some tools that make things more pleasant while we work it out. Also existence of said tools could provide an incentive to go back to that world. If a website is awful, people will browse it with bots and reader modes, a good incentive for people to make their websites not suck.

Your dream wins in a competition of dream quality. But it’s not realistic.

I think of it as the return to 10 blue links. It searches the web, finds stuff and summarizes it so I can decide which links to click. I ignore the narrative it constructs because it’s probably wrong. Which I forgive because it’s the hazmat suit for the internet I’ve always dreamed of.

It gets in the trenches, braves the cookie popups, email signups, inline ads and overly clever web design so I don’t have to. That’s enough to forgive its attempts to create research narratives. But, I hope we can figure out a way to train a heaping spoonful of humility into future models.