There is sooo much hype around this, when the Eon team did almost nothing new compared to the published research. (For context I did my PhD in a fly motor control lab and still follow this field. My colleagues are authors on the research papers Eon used.)
I want to highlight some limitations of the current state of research as well, based on questions that many neuroscientists in this field are struggling with.
- The movement of the legs is not modeled at a fine level, just whether fly is moving forward or turning. This is because we don't have great data on fly leg movements on all these situations, and ventral nerve connectome is still in progress
- By the way the brain connectome still has a lot of errors and needs more proofreading. Also the identity of many neurotransmitters and synaptic strength of connections is unknown. Many are identified through our knowledge of fly genetics which won't translate to humans. In current research, scientists add some finetuning to match some behavior to account for those unknowns.
- There's definitely fly behavior data at the level of making decisions, but not much at level of limb kinematics. Even where data is available, it's unclear how to evaluate the simulated fly against the data. How do you know you got it right?
When I saw the Eon announcement, I was curious how they tackled these challenges. Seems like they didn't. It looks like they forked a few research repositories and vibe coded something to combine them.
I'll give them props for the videos and marketing though, it's crazy to see so many people interested in this research field!!
This is fascinating! Having a really strong video encoder model and then a simpler decoder from that reminds me of the recent D4RT from DeepMind as well: https://d4rt-paper.github.io/
I think we'll see more of these video encoder models in the coming years, they truly seem like magic.
Does streaming compression work if some packets are lost or arrive in a different order? Seems like the compression context may end up different on the encoding/decoding side.. or is that handled somehow?
This is amazing, I just ran it and it works perfectly for my needs!
For context, I recently switched to the KDE window manager (KWin) after a decade of xmonad, to simplify my configuration. KWin supports some tiling but isn't really built for it, so I had some minor annoyances. I ran cortile and it perfectly auto-tiled my windows and allows me to still adjust the sizes with the mouse!
Thank you to the author!
I'd say some default shortcuts conflict with commonly used browser shortcuts, namely ctrl-shift-t and ctrl-shift-r . It's quite easy to configure these, but I found it to be a strange choice for default shortcuts.
I'm not OP, but I've been interested in something like this, but from the perspective of memory systems within oral cultures. I'd love to talk more!
I wonder if you know (and maybe have thoughts about) the arrangement of ancient Cusco, set up to be possible to navigate without any written directions (as the Inca effectively functioned without a writing system).
From Lynn Kelly's Memory Code:
> The Inca turned their major city, Cusco, into a massive memory space, the details of which were documented by the colonising Spanish. Radiating from the Coricancha temple in the centre were over 40 pilgrimage pathways known as ceques. The ceques divided the land into wedge-shaped political, agricultural and irrigation zones, each assigned to a specific kinship group. It is still unclear the degree to which the ceques were physical paths and how much they were purely imagined. To form a city-sized memory space, it does not matter as long as the pathways could be followed in the minds of the users.
I've been thinking about how memory intersects with navigation, and how both of these influence how we interpret the world.
I read her whole retrospective last summer and it was the story that made me feel like I not only had a place as a trans woman in science, but could be ambitious as well!
Retrospective: https://ai.eecs.umich.edu/people/conway/RetrospectiveT.html
It also so charming to see other trans people here on hacker news celebrating Lynn Conway. I didn't know there were so many of us here!
There was a radiolab episode about the other side, interviewing the people advocating for critical theory in debate: https://radiolab.org/podcast/debatable
It's true that it goes against the debate in the moment, but if you zoom out and look at the role of debate within greater society, I think it makes sense to challenge the topics brought up for debate and the whole system that we live in.
I don't know, as a neuroscientist this is very confusing. Blue light generally is linked to arousal, as in this study for instance: https://pubmed.ncbi.nlm.nih.gov/32973462/
I wonder if there is some paradoxical effect if you put in a lot of blue light you get a decrease in arousal? Perhaps it could also be something like repetition suppression, where there is a lot of activity and arousal initially, but this causes a backlash and you get a relaxing effect?
Ahh that's right, I remember using the steam-run wrapper a lot.
If I remember correctly, it would work in 80% of cases and fail on the last 20% (usually if some library is missing or if it has some strange binary installation procedure).
The 20% made it not worth it for me to use as a desktop environment. I still see the value in it for server configurations, though.
I've had a rather similar experience with NicOS a few years ago. I did get comfortable with the language and even contributed a small package! I did like having a reproducible config file for my whole OS setup.
However, everything was just a hassle to get working. Any program that's not in the package repos takes a day or two of fiddling to install. I just lost patience with it and have been on Ubuntu for the past few years.
For the past couple years, I've been using this hacked version of xf86-input-evdev (by Teika Kazura), which allows the use of a space bar for the space and as a "control" modifier if held: https://github.com/lambdaloop/at-home-modifier-evdev
I made a small modification so that if the space key is pressed shortly after a regular character, it just inserts a space immediately. This makes typing feel more natural, with spaces inserted as usual.
I like this better than using the comma as a modifier, as the space bar is nicely positioned on the keyboard to take advantage of your thumbs already.
Day to day, it's such a natural extension that I forget that I have this on. However, this is the first thing I install when I set up a new computer, as otherwise a lot of shortcuts are much more strenuous.
Edit: I just realized that, while editing this, I've been pressing space-backspace (aka ctrl-backspace with this module) to delete whole words. I guess it's fully ingrained now!
It's funny, I was just thinking about this article this morning.
When I first read it over 10 years ago, this quote really struck with me:
> Another personality defect is ego assertion and I'll speak in this case of my own experience. I came from Los Alamos and in the early days I was using a machine in New York at 590 Madison Avenue where we merely rented time. I was still dressing in western clothes, big slash pockets, a bolo and all those things. I vaguely noticed that I was not getting as good service as other people. So I set out to measure. You came in and you waited for your turn; I felt I was not getting a fair deal. I said to myself, ``Why? No Vice President at IBM said, `Give Hamming a bad time'. It is the secretaries at the bottom who are doing this. When a slot appears, they'll rush to find someone to slip in, but they go out and find somebody else. Now, why? I haven't mistreated them.'' Answer, I wasn't dressing the way they felt somebody in that situation should. It came down to just that - I wasn't dressing properly. I had to make the decision - was I going to assert my ego and dress the way I wanted to and have it steadily drain my effort from my professional life, or was I going to appear to conform better? I decided I would make an effort to appear to conform properly. The moment I did, I got much better service. And now, as an old colorful character, I get better service than other people.
> Many a second-rate fellow gets caught up in some little twitting of the system, and carries it through to warfare. He expends his energy in a foolish project. Now you are going to tell me that somebody has to change the system. I agree; somebody's has to. Which do you want to be? The person who changes the system or the person who does first-class science? Which person is it that you want to be? Be clear, when you fight the system and struggle with it, what you are doing, how far to go out of amusement, and how much to waste your effort fighting the system. My advice is to let somebody else do it and you get on with becoming a first-class scientist. Very few of you have the ability to both reform the system and become a first-class scientist.
I tried to live that way for a couple years. Frankly, I think this way of living is unnecessarily restrictive. So what if you get slightly worse service? The clothes you wear and your expressions highlight your history and your culture. By self-censoring yourself, you end up just perpetuating the censorship of other views in the workplace.
This goes double for scientists, as we are rather public facing and have room for wearing nontraditional clothes within our jobs.
This is with a regular Google Workspace plan, independent of the university-wide research plan. The research plan unlimited storage is indeed going away for us as well ( https://itconnect.uw.edu/tools-services-support/software-com... ), which is why we migrated to our own private Google Workspace.
I haven't heard about the google workspace enterprise unlimited storage going away anytime soon, although perhaps you know something I don't?
I work it in a fly neuroscience lab and we use it to store all our electrophysiology and video data. Each person in the lab is storing on average 5TB of data, and the lab as a whole stores 100TB.
The graphical user interface combined with unlimited storage for Google Workspaces is essentially an unbeatable deal. Researchers can upload their data easily through the interface. Any custom solution based on S3 or equivalent would take some time to teach and more time to maintain. Also, we're paying about $200 / month total to store 100TB of data in the cloud, which is hard to beat with other services.
I tried setting up a single account for the whole lab once, but we ran into the above 5M file limit, so we just have individual accounts per researcher and it's mostly fine for now.
Responding to this video, I really liked this take from Aaron Blaise (animator on Beauty and the Beast and Aladdin and director of Brother Bear) on AI in animation.
> To be honest with you, the biggest threat to our industry is not the AI.
I mean when you look at any technology's ever come along in the entertainment industry, the expressiveness, the ability to create new films and new ideas has exploded.
As an end result it has actually created more jobs and budgets have actually gone up.
If you look at the budgets now compared to what they were back in the 90s, they're doubled.
> The biggest threat I think is an age-old threat: it's bad stories and bad management.
That's what's going to get rid of jobs, if you're putting out consistently bad movies under bad management, because no one's going to go see them.
> So the idea is that it doesn't matter what technology you're putting on it. Make sure that the story underneath all of that technology is good, make sure it's engaging, make sure it's something that people want to go to, so they can laugh and cry and be scared and learn something.
As I was told, pretty much any PhD graduate can get a postdoc somewhere. PhD -> postdoc transition is really about whether you want to stay in academia at all. From my rough impressions, maybe about 10% of the grad students I know are thinking of doing a postdoc, which would be 10% * 15% = 1.5% of PhDs making it to faculty. That roughly matches the statistic you have.
I think the 15% of postdocs becoming faculty statistic really sticks with me, because these are already preselected for wanting to become faculty and even within that pool, only 15% can make it. It's quite demoralizing.
Yes absolutely this! This is 100% a thing in the US as well. Moving around between undergrad and PhD has wrecked my social life. Many of the students in my program are not doing a postdoc because they don't want to move out of Seattle. And I mean, why should they?
In the US, the only places where you could reasonably stay in the same place and switch institutions are LA (and surrounding area), Boston, and New York. Anywhere else you're switching states and potentially leaving friends, family, and loved ones behind. It's really hard.
I want to highlight some limitations of the current state of research as well, based on questions that many neuroscientists in this field are struggling with.
- The movement of the legs is not modeled at a fine level, just whether fly is moving forward or turning. This is because we don't have great data on fly leg movements on all these situations, and ventral nerve connectome is still in progress
- By the way the brain connectome still has a lot of errors and needs more proofreading. Also the identity of many neurotransmitters and synaptic strength of connections is unknown. Many are identified through our knowledge of fly genetics which won't translate to humans. In current research, scientists add some finetuning to match some behavior to account for those unknowns.
- There's definitely fly behavior data at the level of making decisions, but not much at level of limb kinematics. Even where data is available, it's unclear how to evaluate the simulated fly against the data. How do you know you got it right?
When I saw the Eon announcement, I was curious how they tackled these challenges. Seems like they didn't. It looks like they forked a few research repositories and vibe coded something to combine them.
I'll give them props for the videos and marketing though, it's crazy to see so many people interested in this research field!!