benliong78

Uruky sounds like the google that we used to have: one that's just good because it returns relevant results, not because of ecosystem lock-in.

That being said, can't help noticing the [NO-AI] at the beginning of the post, and I wonder whether it would help adaption if you also make the search engine available to agents for paid customers.

that animation of Altman with a thousand faces is oddly unsettling. Good job, new yorker.

A email thread, to me, is much preferable to any IM, unless participants knows to send those IMs in a thread as well.

All the context are in there in the rawest of forms, you a run through them with your eyeballs or have your tools do the summarisation there and then. Most of the IMs I received didn't even quote the original message they were responding to, and I end up spending time jumping up and down the channel or group to get the whole context. Not to mention folks sending me link to a slack, which, depending on the mood of the almighty slack god, can or cannot be opened in the app / current slack session.

But you do you. :)

That’s fantastic. In your opinion what are some of the best books / resources you use to have this kind of understanding of LLM and the underlying deep learning algorithm?

I’m building a on-device AI App that supports a good variety of open-weight LLM Models. It’s mostly so that I can gain an intuition on what goes on when we do inferencing, but my longer term plan is

1. Support on-device RAG to allow chatting with your own documents on mobile offline 2. Support MCP on-device, taking advantage of information that’s (only) available on your phone, like calendar events, health data, etc. These shouldn’t need to be anywhere but on-device. 3. Allow on-device AI to use shortcuts(?)

I think most of the functionality are well served on desktop front with Ollama and LM Studio, but moving these functionality to mobile offers a great learning opportunity.

Sadly that may not be far from the truth.

China folks are more hungry than most in the west, with the exception of Silicon Valley and New York, and China isn't "burdened" by quality of life, and work life balance, which isn't healthy, but that's one of their edge.

Even for U.S. AI Startups, asian makes up a majority of the research + technical staff. I think the incentive to go into PhD in the west has been gradually declining for years, while PhD programmes have been flooded with Chinese Students. I think it's something that the U.S. government needs to address to try to remain competitive in the longer term.

Manufacturing is something that the west actively gave up and migrated the skillset to Asia, notably China and Taiwan.

I think while technically they differ mostly only in terms of the training materials thrown into it, the outcome is that each model is good at something and bad at others, just like human being. Soon you'll need standardized tests and HR department to evaluate individual LLM performance. :)

Sigh. But honestly, it just felt like people getting used to China doing outrageous things like these to the point where it's no longer interesting for a comment.

dude, that’s freaking nuts. It’s amazing and terrifying at the same time.

I don’t know man. I mean, google indexes the open web to create a search engine and we all seem to be fine with it. I do draw the line that if you want to build a LLM, there’s the internet for you, train your own damn model and stop piggybacking on someone else’s, or in China’s case, at least don’t get caught.

Remember, "A little bit beyond Hello World" in Carmack's mind is Video Decompression code. Brings a little bit of perspective.

We in Hong Kong are all assuming that TikTok is CCP-owned at this point.

Young children are obsessed about TikTok, even in Hong Kong, and parents are just too happy to let their kids melt in front of a mobile screen all day just to get some rest.

But then there's also Zero chance that the CCP is telling them to leave Hong Kong. TikTok imho is a spying and influence tool for the CCP anyway.

This move feels very very weird.

You're right that correlation does not equal causation, but it doesn't have to be for those folks to not care.

Having lived in Singapore, Hong Kong (current) and China, I can tell you that in Singapore's case, folks are aware that they live under a semi-dictatorship, that their government controls all media (radio, tv, everything is state-owned). But they always say "but I guess the government is doing a good job".

One has nothing to do with the other. The government can do a good job without the dictatorship, but it's the way folks rationalize their helplessness, their sense of lack of control over their fate.

This is also why we in Hong Kong ARE fighting against these kind of controls, probably in vein, but I for one do NOT want to resign in rationalizing for the government's over-control without a fight.

It looks like something that's implemented in China, and we don't have problem with Ice drivers spitefully taking out multiple spots. What we have problem with is Ice drivers parking in Tesla Charging Spots (and this applies to other EV spots as well), just because they need a space to park. This system would solve that problem pretty well.

I drive a BYD in Hong Kong and we see Tesla drivers taking up charging spots just because it's marked "EV only", but they can't use BYD chargers <shrug>

Kinda wish you named it: 'Robot or not' https://www.theincomparable.com/robot/

Audible has something similar: they require that you use up all your credits before cancellation is even possible, so that is obstacle #1. Then they ask for reason and sub-reason, and give you either a $20 coupon if you say the reason is because it's too expensive, or a 3 months praise on your account if you say you can't catch up to one book a month.

Pretty clever. But having to choose 6 books immediately at the point of cancellation sucks

Thanks for the recommendation. Fantastic looking software that provides a far better visualization of the repo than anything else I have used. Bravo.

I use CLI for almost everything but sometimes it helps to have a visual representation of the repo.

Living in another crowded city, Hong Kong, perhaps I can offer some perspective: We have tons of elevated tools highways and bypasses aimed at reducing traffic, the problem is all these elevated roads needs to go around a lot of tall buildings. Going underground with subway already provides us with better solution to a lot more of our traffic problem. What the boring company's proposing isn't all that different from the subway solution (we already have subway going three or four level deep), it simply allows cars to transported via subway.

Literally

I'd say the same as someone who lives in Hong Kong and travels to Shenzhen with some frequency.

With one caveat, those folks selling parts definitely treats me with a lot less friendliness than they do with foreign buyers, especially when you're buying just one.

There are also varying level of quality when buying parts, especially batteries and screens. It's hard for repair shops in the area to charge higher prices for better parts, so there's definitely a huge market for screens that's slightly mis-calibrated, has a non-noticeable dead pixel and such. Customers in the area generally care less about those kind of quality.

Wow that's a lot of learning.

By MIT self driving course are you referring to http://selfdrivingcars.mit.edu/ ?

Swift is simply a safer language to use, safer as in it tries to prevent you from shooting yourself in the foot. And at this point the framework support has pretty much caught up. So my vote is for Swift on new projects.

And stick with Swift 3.0

That's a fair point, and I've lived in NYC briefly to know that it is indeed annoying to see the driver do circling around in traffic to get exactly to the door.

You can use AI to solve it, or you can give the user power to tell the driver that he/she is willing to work a little bit. A simple button may have been suffice

iOS gets close to what you want. Apps have to ask for specific permission for specific reasons (location, background download, etc), otherwise it just get killed after a few minutes.

Security is always at odds with convenience. It's easy to require a pop up to ask for permission every time and have user getting so used to these pop ups that they just automatically tap "yes"

So the question is valid: what benefit does one get using Rush over Swift on developing for Mac? Rust may be a better language overall, while still be an inferior choice in this context of Mac OS X Native app development.

What could possibly go wrong, really.

Some would argue that Gathering training data is beneficial to users in the long term, and the more aggressive auto-manufacturers go about it the better the cars will be to us users.

I agree that Tesla is using the term "autopilot" opportunistically to say the least. Whether the term accurately represents the feature they want to promote (imho technically it does, but more on that later) is becoming less relevant.

But I'd like to bring counter-argument to your points (a) and (b): taking away the autopilot feature from both airplanes and cars, an airplane pilot still has to undergo substantially more training than a car driver does, so imho the increased training for pilots compared to vehicle drivers is largely coming from the difficulty in controlling the machine and the more deadly consequence of control-loss to the driver / pilot and to the others.

That being said, the term autopilot is being used opportunistically. There's no way Tesla will back away from the term now, so the next best thing is to warn the user, in every step of the way, that "YOU HAVE TO KEEP YOUR EYES ON THE ROAD IDIOT", to the point that if they can detect your hands not on the wheel a warning sounds.

I do believe the Tesla Autopilot and the autonomous mode in the future will save lives, what we're experiencing is the growing pain till we get there.