> Nearly every species of scorpion intensely fluoresces under UV light. […] Scorpions have photoreceptors in their tails, separate from their eyes. […] It is hypothesized that a scorpion uses this fluorescence to tell whether any bit of its body is left exposed from its hiding place. Its tail “looks” down at its body, and if it sees its own fluorescence, it knows it is exposed to light, and in danger.
And a special call-out to the “Andean Cock-on-a-Rock” :), see a photo in the article.
It must be more general than that, otherwise the cells wouldn’t be able to repair their area if the damage came from the wrong direction (repair is not center-out).
The model generally learns to generate each pixel from its surroundings, even if the surroundings are partially missing.
There's hidden state in the model which presumably it uses to communicate position, ie there's the 3 colors but then a bunch of other channels that the model can use how it wants.
To me, it is intriguing as a toy model for how cells are able to grow into complex tissue and organisms based only on local information, and how they are able to repair and recover harmed tissue.
Of course, this is as close to cells, as neurons from neural networks are to real neurons. And I have no idea what it could be applied to (inpainting/outpainting?), but it’s interesting as exploratory research.
Oh scientists are leaving science in droves, certainly. Often becoming sales-people for deep-tech companies, which is rather sad.
This is the most recent shock, and probably the biggest one, but academia has been becoming toxically metrics-driven, authoritative and political for a long while, weirdly more than in industry.
It has nothing to do with scientists of course, they are the last ones that would want this. It's a never-ending squeeze from the top.
And also the fact that so many students were pushed to study pure sciences, which is great in principle, but some of these degrees only prepare you to stay in university as an academic, and there's only so much budget for that.
In EU there are laws that force universities to give researchers a permanent contract after a couple years. The result? Everyone gets fired every couple of years. In certain fields, this implies changing country every couple of years.
Not that the university is paying much anyway, often the opposite: the researcher gets their own grant and they are forced to pay a cut to the host university, or to their group leader. It can get rather feudal.
The actual law is more that you need an objective reason for a fixed-term contract in any sector. A genuine project (such as the completion of a PhD) is an acceptable reason. The availability of funding is not.
In practice, it has been accepted that postdocs can have fixed-term contracts, because it's a trainee position. Similarly, an assistant professor can have a fixed-term contract before tenure. Both of those are in some sense against the spirit of the law, but the legal system tends to favor consistency and reasonable outcomes over strict adherence to the law.
European universities have more postdocs than American universities, because there is more research funding available. But then there are fewer faculty positions for those postdocs, as the universities themselves are not so well funded. That creates a constant stream of researchers looking for other opportunities, which American universities used to take advantage of.
Universities tend to operate strictly on a budget, because they only have limited discretionary funds. While a business may choose to buy things it believes it needs, because it expects to make money in the future, a university generally needs to secure the funding first. If you are a researcher, you don't get an office, a laptop, and some lab space simply because you need them to do your job. You may get them if an external funder explicitly chooses to pay for them.
I had some visibility into the funding of Finnish universities during the transition to the current system. Under the old system, core funding was more generous. Each university allocated the resources between various units and individual professors, which involved a lot of politics. If someone was particularly successful in obtaining external funding, they might not have enough office/lab space for all the people they could otherwise hire.
The funding model gradually changed to address issues like that. Departments had to pay internal rents to the university for the facilities they used. The government started allocating some of the core funds according the success each university had in obtaining external funding. And at some point, they moved most of that money from core funding to grant overheads.
As far as I remember, from when I was closer to academia, in NL postdocs had to be offered a permanent contract after 3 temporary contracts, with a maximum of 1 year per temporary contract, or something like that. I believe this wasn’t exclusive to postdocs and it is a general law for most professions.
In recent years in Spain they aggressively decreased that threshold to the point where most employees need to have permanent contracts. Interestingly it has led to significant growth because, among others factors, it has increased consumer confidence, and it has been a much smaller burden on companies than expected.
Perhaps the term “permanent” contract is confusing to some. It’s not in the sense of a functionary or tenure, where you virtually have a job for life unless there are extreme circumstances. A permanent contract is an indefinite contract, one without a specific end, where firing you needs to be properly justified, but you can be fired, certainly.
An employee is not a pure machine that converts money/time into results linearly.
It bewilders management, because there's a very significant overhead involved in making sure an employee is properly synced on what needs to be done, making sure they are content and productive, and managing the administrative logistics around them. Even disregarding the work of management, in a flat team the communication overhead that each member adds can also be significant and non-linear.
Generally, adding people adds a lot of complexity and inefficiency to an organization, and if you can do something without more people that's usually a lot better. It depends on the role of course, but in many jobs now an employee that is not fully dedicated can be a net-negative. The same can be said of employees that are not very experienced or competent.
This is why there's a significant crisis in early-career employment. More generally, it's also why we have a large fraction of population feeling like they cannot get a decent job, while many companies are simultaneously struggling to find the employees they actually need for a reasonable salary.
I work on 90% contract, meaning I get cca 10 weeks of paid vacations yearly, and its usual 5 days a week workload. Net income hit is somewhere around 6-7% of salary.
There is maybe tiny overhead, but there is also more efficiency during time I am actually in, especially in slow moving processes. Plus QoL improvement is massive for me, as an adventurer, mountain lover and first and foremost a parent of 2 young kids.
People are scared these days to look for new job, its same as it was in 2008 in many regards (I personally went in opposite direction during that time despite many people warning me against, and actively started consulting and soon after then relocated to Switzerland), but our lives are short.
Do you want to end up regretting working too much for some empty goals of others, which usually #1 regret of dying people? I sure as hell won't be in that category, company performances, insecure egos of control freaks in management and other bullshit be damned, they are not meaningful part of any life well lived.
Ah yes, of course, that's not what I meant. I would count you as fully dedicated, what you are describing is not too rare in EU in some professions. And I'd say that getting long vacations is quite a different dynamic than working part-time on a weekly basis.
I was referring to the commenters talking about working 2/3 days a week. In the Netherlands 4 days a week is also becoming the norm, which I'm not a big fan of but it's not all that bad either, actual productivity doesn't change that much in practice.
I just mean that at some point, if you are not actually focused on your job, you end up creating more work than you deliver, or at least not enough of a surplus to justify a salary. So it's not surprising that managers are averse to reducing hours and salary linearly, the impact is not linear.
I really think this depends on the job. As soon as I read your initial comment, I thought about locums in medicine, people who float for as little as a weekend at a time, and as a little as once a quarter at any particular hospital. And the entire hospital industry has been built around them at least in the western United States. They’re clearly contributing something.
I think there are jobs where you need lots of context and there are jobs where other things are more important.
> Pricing for both models is $10 per million input tokens and $50 per million output tokens.
The step-up in intelligence looks massive (we'll see in practice), but the price is getting to a point where it's making me question if it's even worth giving it a try.
Good competitors will probably be out soon, which should level the playing field. I am more excited about that, just the fact that they showed that such an improvement is possible. I'm okay waiting a bit longer for this to become attainable for plebs like me.
Models are getting better, but there's a negative change in terms of "productivity" per dollar. Yeah, I can throw 5 sub-agents at the problem, but the cost is getting significantly higher. And yes, I can crank out the solution much faster, but again, at some point that cost will be hard to justify. And it doesn't matter if the cost is subsidized by a provider, if it's paid by your company, or from your pocket. We are slowly reaching a point where the cost will be too high to justify the gains.
Sadly this does not seem to be the case here: if you read the announcement entirely, they include a "cost per task" metric which basically continues the trend of their previous models. So yes, tasks will cost you more, but results will be better - allegedly.
after playing for the three days - yeah, that turned out to be the case. Fable was more precise, did more tests and cost much more than 2x for the same task.
For some tasks though Opus was performing poorly and Fable managed to do them well on a first try.
I'm not sure how it might be with Fable in practice, but we are already not that far away from AI costing as much as a full-time professional, faster in some ways but considerably less independent.
Perhaps not that close to US salaries, but those are inflated to hell. Worldwide senior engineers and scientists have salaries just about an order of magnitude away from AI subscriptions that you can use most of the day every day.
Prices have fallen dramatically over the last few years. It’s just that our standards have increased because we are using AI in ways that were not possible with worse models. But for the same level of “intelligence” as we had a couple years ago, the prices are so much lower.
This kind of immediate-mode rendering is quite standard for TUIs. Although immediate-mode rendering tends to be significantly simpler and use less memory than retained-mode rendering, at the cost of some redundant computation. So I am not sure if this is the reason for the bloat.
It’s possible that it doesn’t play well with JS garbage collection, since it recreates the whole UI structure for every frame (which tends to not to be an issue in the languages immediate-mode is usually employed).
But yes it’s a bit more akin to game renderings than web rendering. Which can be totally fine if done well.
I haven't tried to make a TUI admittedly, but double buffering is the oldest technique on the planet. A TUI doesn't even need to pay the cost of a lot of pixels since its effective resolution is much lower
> Nearly every species of scorpion intensely fluoresces under UV light. […] Scorpions have photoreceptors in their tails, separate from their eyes. […] It is hypothesized that a scorpion uses this fluorescence to tell whether any bit of its body is left exposed from its hiding place. Its tail “looks” down at its body, and if it sees its own fluorescence, it knows it is exposed to light, and in danger.
And a special call-out to the “Andean Cock-on-a-Rock” :), see a photo in the article.
reply