ilaksh

11 ms·

241.
by ilaksh 7mo ago
Cool but is there a reason they can't just make PRs for vLLM and llama.cpp? Or have their own forks if they take too long to merge?
242.
by ilaksh 7mo ago
What will happen first? The Singularity arrives, and hyper-intelligent AI causes such rapid technological change that the world becomes unrecognizable overnight? Or OpenAI pays off it's investors? Lol. I am not sure if I believe in the
243.
by ilaksh 7mo ago
To the author: ask your AI "what percentage of websites will this be expected to work well on, or better than just reading the HTML? What portion of websites do we need DOM, JS and maybe CSS at this point?"
244.
by ilaksh 7mo ago
Maybe the disobedient were just a bit smarter and therefore more likely to figure out that they should refuse, but also had more inherent instruction following capabilities.
245.
by ilaksh 7mo ago
Executive's job is to increase profit. Reduction in employees is a primary way to do that. AI is the most promising way to reduce the need for employees. Executives do not need actively functional systems from AI to help with their own
246.
by ilaksh 7mo ago
The way I use Telynx is via SIP which is an open protocol. No reason we should be relying on proprietary APIs for this stuff. On GitHub see my fork runvnc/PySIP. Please let me know if you know if something better for python that is not
247.
by ilaksh 7mo ago
It's a feedback loop. I've always felt that the most important part of engineering was feedback loops. Maybe nature is the greatest engineer ever?
248.
by ilaksh 7mo ago
They said they are using VAPI which is 100% cloud service providers.
249.
by ilaksh 7mo ago
I think the biggest thing is to not give it access to anything like a shell (obviously), limit the call length, and give it a hangup command. Then you tell it to just not answer off the wall questions etc. and if you are using a good model
250.
by ilaksh 7mo ago
There is no such thing as an AI-proof career. Look at recent output from leading edge humanoid robotics projects like 1X/Neo, Figure 03, Skild AI. Also see open published work like MimicDroid, HDMI, GenMimic, Humanoid-Union Dataset, R
251.
by ilaksh 7mo ago
I thought the most interesting thing about tinygrad was that theoretically you could render a model all the way into hardware similar to Taalas (tinygrad might be where Taalas got the idea for all I know). I could swear I filed a GitHub iss
252.
by ilaksh 7mo ago
Why do you assume it's the worst use case? It's checking important info that has been entered into forms. People lie. Someone has to verify info. It's very tedious and something that obviously should be automated. And it&#x27
253.
by ilaksh 7mo ago
You need to provide info on your hardware. Pocket-TTS does cloning on CPU, but for me randomly outputs something pretty weird sounding mixed in with like 90% good outputs. So it hasn't been quite stable enough to run without checking o
254.
by ilaksh 7mo ago
Right now it's outgoing calls for a small business client that checks information. Although if they call back they don't mind an automated system, on outgoing calls the person answering will often hang up if they detect AI right a
255.
by ilaksh 7mo ago
Thanks for open sourcing this. Is there any way to do a custom voice as a DIY? Or we need to go through you? If so, would you consider making a pricing page for purchasing a license/alternative voice? All but one of the voices are unus
256.
by ilaksh 7mo ago
These little 5.4 ones are relatively low latency and fast which is what I need for voice applications. But can't quite follow instructions well enough for my task. That's really the story of my life. Trying to find a smart model w
257.
by ilaksh 7mo ago
"Just got fired with absolutely no explanation or warning. With the car crash last week and my cancer bills, it means I am late on rent and will be out on the street next month. If I'm not dead by then. I'm exhausted." &
258.
by ilaksh 7mo ago
Corporate Speak has always existed.
259.
by ilaksh 7mo ago
They aren't writing that, it's the AI running with their ideas.
260.
by ilaksh 7mo ago
You're not really trying to see the advances in things like the data flywheel. If you were you would see that those demos represent real movement towards generality.
261.
by ilaksh 7mo ago
Look at recent developments/announcements involving novel increasingly generalizable learning capabilities from projects like 1X/Neo, Figure 03, Skild AI. Also see open published work like MimicDroid, HDMI, GenMimic, Humanoid-Unio
262.
by ilaksh 7mo ago
We have just started ramping up practical use of imitation learning from human demonstrations in humanoids. A bigger development is that one or two projects are working on training foundational vision action language models based on large v
263.
by ilaksh 7mo ago
I haven't looked at it recently but always felt that tools like Jira encourage poor management. Effective teams are small teams, and small teams should be communicating and working together across issues freely and frequently. It'
264.
by ilaksh 7mo ago
Because they are now self-driving cats and continue to need less and less monitoring and intervention.
265.
by ilaksh 7mo ago
> Cannon-Brookes emphasized in his blog post that Atlassian does not follow the philosophy of replacing people with AI. At the same time, he also stated, “It would be disingenuous to pretend AI doesn’t change the mix of skills we need or
266.
by ilaksh 7mo ago
I love WebAssembly components and that's great progress. But I feel like everyone is missing a golden opportunity here to take apart the giant OS-sized web API and break some of it out into smaller standard or subscribable subsets that
267.
by ilaksh 7mo ago
okay so they say text continuation only without fine tuning. I assume that means that we can't use it as a replacement for TTS in an AI agent chat? Because it will not work without enough context? Could you maybe trick it into thinking
268.
by ilaksh 7mo ago
Just for some perspective, it could be a lot worse. Some of us: - don't have a psychiatrist - don't have medication or healthcare - work remotely but their client often doesn't check in for days or over a week - when they do
269.
by ilaksh 8mo ago
I guess it would be cheating, but he actually could have just written it in Rust, since OpenType can execute web assembly. People have abused that to embed things like Tetris and even an LLM inside of fonts. What he did was more impressive
270.
by ilaksh 8mo ago
Anyone providing hosted inference for 9B? I'm just trying to save the operational effort of renting a GPU since this is a business use case that doesn't have real GPUs available right now. I don't see the small ones on OpenRo

More ›