Robin Glauser

@garrytan I actually tried once to get gemini 1 to convert music videos into the ultrastar format and it almost worked (The formatting and speech recognition was ok, better if I supplied the lyrics, but the timing was a bit off). I think a fine tuned model might make this work.

As a karaoke fan, out of pure curiosity I wonder if you all have found any open source that can auto convert any music video into a karaoke experience using AI yet? (Fun fact: my senior project in college was something that converted any MP3 into a Dance Dance Revolution song)

@burkov Yesterday, my account was mistakenly deactivated as well and then reactivated an hour later. However, my Plus subscription is now missing. If they don't restore it, I'll likely switch to using Gemini more and won't resubscribe unless they release a new model.

OpenAI deleted my account without explanation. The email they sent said "OpenAI's Usage Policies restrict the use of our services in a number of areas. We have identified ongoing activity in your account that is not permitted under our policies." I only used ChatGPT for coding, and lately not so much, because the o3 model was unavailable most of the time. The only reason I can see is my online posts about Altman's nose. It's the lowest thing I've ever witnessed from a tech company.

@milesabovetech Seems to be this file here: thingiverse.com/thing:6783386/… The trick is to add a gcode pause when it get's to the upper layers, then change the filament to black, so you get a nice contrast and don't need to color it with a pen.

They say 3D Printers are useless, I beg to differ…

It's a big day for Claude's PDF capabilities. We're rolling out visual PDF support across claude dot ai and the Anthropic API. Let me explain:

As we’ve been harping on about for many many years: ALL ALGORITHMIC DECISION MAKING MUST HAVE MEANINGFUL HUMAN RECOURSE.

@kamathsblog CleanShotX for Screenshot: cleanshot.com Transmit for Transfering Files: panic.com/transmit/ Iterm2: iterm2.com Rectangle for Window Management: rectangleapp.com Parallels if you feel nostalgic for Windows or Linux: parallels.com

Hello #tech Twitter, I have never used a MacBook before and I just ordered one. Looking for recommendations for must-have software, productivity tips and tricks and accessories that I must get.

@lightxvision I did that in my last year of apprenticeship. I was so happy when I finally got X running with Bumblebee so I could use the NVIDIA and Intel GPU.

There should be a university cs program where all freshmen are given a cheap laptop with no OS on it, have to install Linux on it themselves, then do all of their programming work on that with no web browser for the first two semesters

@willdepue And keep the phone screen on and unlocked with the setting open until you have connected.

@willdepue My trick is to open the hotspot settings on the phone, then it appears right away on my mac or other devices.

i want the apple engineer responsible for it taking ~10 minutes to connect to my phone hotspot from my mac to know that you bring great misery to my life and have sucked a great deal of productivity from the world. please. please fix this.

@asallen Having to use multiple providers like outlook, gmail and finding a client that has support for all of them. Also customizability and it would be nice if we could have a modular calendar app that you could extend easily.

What's your biggest gripe about your calendar?

@shanselman Have you heard of UTM? getutm.app @parallels on iPad would be awesome though.

But mine on Facebook said let the iPad run virtual machines and it would solve 80% of the problems with 20% of the effort. Absolute genius idea.

if your data is stored in a database that a company can freely read and access (i.e. not end-to-end encrypted), the company will eventually update their ToS so they can use your data for AI training — the incentives are too strong to resist

@EmileAndHisBots That's the question I often ask myself and I get to the same answer. Not sure if we should to replicate all our current processes with humanoids. For some processes it will be much better to rethink them and use other technologies to improve them.

7/ (Yes, the common argument is that spaces are optimised for humans so the robot needs to adapt for that - I don't buy it, 99% of the time rearranging the space is cheaper than any tech solution)

1/ So, looks like humanoids are on the menu again! I've been working on "humanoids" (NAO and Pepper) for over a decade, so: why humanoids? A 🧵:

The Windows 11 Start Menu is comically bad. This machine has a $1600 Core i9 CPU and 128 GB of RAM and this is the performance I often get. What is going on in Redmond?

@audrow And here is a video about it m.youtube.com/watch?v=0LfJ3M…

@audrow In the Tesla Optimus Video, you can see that they implemented touch sensing at 1:14 m.youtube.com/watch?v=cpraXa…

Is touch sensing important? I burned the tips of three fingers a few days ago, and I haven't been able to feel much with them since. I'm dropping more things. I feel clumsier without good touch sensors. Robots have been doing some amazing things in object manipulation, and most of it to my knowledge is without much for touch sensing. I wonder how much better robots will be with good touch sensors.

@audrow This is the study pubmed.ncbi.nlm.nih.gov/19352402/

My three laws of robotics. The companies that I have founded have sold 50+ million real robots to real customers; research robots, home cleaning, nuclear power plant inspection, military ground robots, upper body humanoids in factories, and now at Robust.AI, intelligent carts in warehouses and factories. I'm aiming to get to 9 digits of robots rather than a mere 8, and in thinking about that I've reviewed my own implicit three laws of robotics that have guided me, and now have written them down explicitly. Here they are: 1. The visual appearance of a robot makes a promise about what it can do and how smart it is. It needs to deliver or slightly over deliver on that promise or it will not be accepted. 2. When robots and people coexist in the same spaces, the robots must not take away from people’s agency, particularly when the robots are failing, as inevitably they will at times. 3. Technologies for robots need 10+ years of steady improvement beyond lab demos of the target tasks to mature to low cost and to have their limitations characterized well enough that they can deliver 99.9% of the time. Every 10 more years gets another 9 in reliability.

@coreylynch Is the speech-to-text and text-to-speech running on edge on the robot?

We are now having full conversations with Figure 01, thanks to our partnership with OpenAI. Our robot can: - describe its visual experience - plan future actions - reflect on its memory - explain its reasoning verbally Technical deep-dive 🧵: