aliljet

I genuinely wonder if this kind of use case is going to completely upend UX testing. There's something dynamic, perhaps poorly so, that would change the way prescriptive, painfully locked UX tests work...

I'm curious about whether anyone is running this locally using ollama?

This is fantastic news. I've been using Qwen2.5-Coder-32B-Instruct with Ollama locally and it's honestly such a breathe of fresh air. I wonder if any of you have had a moment to try this newer context length locally?

BTW, I fail to effectively run this on my 2080 ti, I've just loaded up the machine with classic RAM. It's not going to win any races, but as they say, it's not the speed that matter, it's the quality of the effort.

That's odd, attempting a prompt fails because auth isn't working.

Does this work to assess each individual'd actual caffeine half life? Every individual is metabolizing caffeine at a different rate...

And this is happening RIGHT as a new potential leader is emerging in Llama 3.1. I'm really curious about how this is going to match up on the leaderboards...

Count me marginally confused. Has T-Mobile already increased prices? Who is presently engaged in suing them for this breach of contract?

Is there a reason you're not trying to port this into an even more stack agnostic world without CUDA?

how would you propose this might work?

Who wants to do this in Kirkland, WA?

I've long had a bet that these posts contain lurking single founders looking for potential co-founders. So, if you are on the prowl for a jack-of-all engineering founder, shoot me a note. I'd love to see if you might be all-in on what I've got going on or vice versa. Here's building awesome stuff with awesome teams.

SEEKING PROJECTS: I'm a pretty senior engineer at a series-a funded startup (BeeKeeperAI, we're enabling untrusted parties to work together with fancy confidential computing) having fun with golang, python, and rust. I emerged from UCSF having built and published in machine learning space. Broadly, I'm a guy who likes to build things and I'd love to help you! pav.gup at gmail dot com.

This is why terminating an employee has to be swift and complete. It's just too risky to bet otherwise.

This is incredibly interesting. Why wasn't this seen as fruitful over the last decade?

I've been following this space very very closely and the killer feature would be to be able to generate these full featured videos for longer than a few seconds with consistently shaped "characters" (e.g., flowers, and grass, and houses, and cars, actors, etc.). Right now, it's not clear to me that this is achieving that objective. This feels like it could be great to create short GIFs, but at what cost?

To be clear, this remains wicked, wicked, wicked exciting.

At this stage, I wonder if @sama and @gdb could just form ReOpenedAI? They could at least forgoe to pretense of the work being for some kind of greater good than profit.

This is such an odd circumstance. The OAG is not saying each of the management firms is actively colluding, but is accidentally colluding by agreeing to use RealPage pricing. In practice, this looks like RealPage is simply collecting rent data to determine what pricing looks like in the market.

Yes, these RealPage estimates seem to be used, but if these estimates were wrong, the large landlords would simply fill up their competitor's units and would create an incentive to create more competitor units.

Perhaps the DC governemnt would do better to simply build it's own brand of low-cost housing.

How recently was this LLM seeded with data? In the context of golang, this is easily a generation ahead of GPT-4.

This is so damn cool. How do you discover services in this environment (e.g., how does that weather balloon announce it has an image?)

Could this be because of some of the unionization underway?

I built a house myself.

I'm so thoroughly confused about why AMD wouldn't be falling over themselves to enable geohot and his followers to build an alternative to CUDA and NVIDIA. This feels like a conversation that geohot is attempting with feckless product and software managers who certainly can't make bold decisions. Has the CEO of AMD effectively spoken about this problem?

Robotics is such a lovely world. For the hobbyists, I'm really curious about how to get the home-brew robotics training lab working at home? The last time I reviewd this, you'd spend thousands of dollars just to get a reasonable robotic arm. Right now, if work is relegated to the rich research departments of mega corporations, this certainly doesn't seem more interesting than a corporate press release..

This is wicked interesting. This may sound ridiculous, but I'm curious, how do you get a satellite to space? I can't just get a quick Lyft Line ride to leo, can I? :)

I'm really curious about why you would want to use FastAPI over Django Rest Framework. Are there signficant advantages to FastAPI?

I really really really want to play with robotics at home, but I've discovered it's super expensive to get things like robotic arms at home. Are there robotics platforms that are DIY ready? I really just want to teach a robotic arm to make my coffee in the morning. And I'll apply my life's effort to achieving this only if the cost of the robotics platform is stupid low.

Yes. I know that's absurd. :)

Suppose I had not introduced my comment as being primarily generated by an LLM, how would you have policed this or reacted? And to be clear, my comment is doing quite poorly which seems to show the system is working, but I wonder if the summary, judged exclusively on the merits of its content, would have done as poorly? In the absence of my own personal time and alternate summaries in the comments at the time I posted, I found the two paragraphs from this LLM to be pretty damn good.

And yes, I understand the concern about machines batting around ideas with humans, but it seems like up and down votes are working, admittedly to my disadvantage today.

Yep. You're right. It's really just nav blocking that's killing the drones. I wonder if there's a way to have these drones operate without a need for GPS. Instead, use cameras to read terrain and map that against the known terrain...

I'm a little baffled by this, 'Russia has deployed big ew systems for every 10km stretch of the front. They are largely aimed at drones. And they work: Ukraine is losing a staggering 10,000 drones per month, says the report, amounting to over 300 losses daily. ew is responsible for around half of those losses, a Ukrainian official tells The Economist.'

What Electronic Warfare is going to pull 5,000 drones a month out of the sky?

What did this cost JPMorgan? From the outside looking in, it seems like they walked away with assets and little to no liabilities.