lamename

Some would pursue less lucrative passions. Some would spiral into lazy Sunday behavior every day.

Good writing is hard, but it's also very easy to just regurgitate patterns of nonsense buzzwords you see everywhere else. I've seen many people hear that and just nod their head. Sadly it convinces some.

Sure. As long as all the other fraudsters and science deniers that were just elected or waiting to be appointed go too

I agree that the selection for memorization is high, and I've worked with many neuroscientists who cared more about biological "stamp collecting" than understanding systems.

But in my experience neuroscientists have to have a solid level of systems thinking to succeed in the field. There are too many factors, related disciplines (from physics to sociology), and levels of analysis to be closed off.

As a neuroscientist, yes. But this is true of most science & medicine news written for a general audience. I have to tell my parents to stop reading "Chocolate is a superfood / Chocolate causes cancer" articles every year.

Uncertainty isn't good for engagement, even if it's correct.

These are all true for me, but you know what's worse? In the office I have to hear everyone's meetings all the time. People feel compelled to tap me on the shoulder and interrupt the work that was supposedly so important to get done this sprint.

And you know what? In office, half the team's on video anyway because why hire locally when you can hire nationally or internationally for the best people?

I'll take a few video calls, then some actual privacy for 5 minutes to decompress, so I can actually focus for more than 45 minutes to do my work. The tradeoff is still worth it for me.

Zero Interest Rate Phenomenon

https://www.businessinsider.com/zirp-end-of-cushy-big-tech-j...

Maybe I missed it but I don't see that explicitly in the article. I see someone who learns for fun and for their growth.

That being said, yes people can take it to the extreme just to put something on their resume...working on unrelated nonsense at the expense of the team. I've been on the receiving end of this and can't stand this immature behavior.

In fact, the two times I've experience it, it was the least concretely useful person on the team (but not a junior). When their BS is called out they ended up leaving one way or another instead of changing behavior.

This is different from someone who loves learning and applies broadly learned lessons in day to day work.

Posted by Chollet himself:

> I don't think people really appreciate how simple ARC-AGI-1 was, and what solving it really means. It was designed as the simplest, most basic assessment of fluid intelligence possible. Failure to pass signifies a near-total inability to adapt or problem-solve in unfamiliar situations.

> Passing it means your system exhibits non-zero fluid intelligence -- you're finally looking at something that isn't pure memorized skill. But it says rather little about how intelligent your system is, or how close to human intelligence it is.

https://bsky.app/profile/fchollet.bsky.social/post/3les3izgd...

I think you're right in the long run, but in the short run all it takes is largely the non-experts and business people to _think_ they can make mass layoffs and "hire" AI for a significant number of people to lose out on the industry.

They will certainly change the output, as we've already seen, likely not for the better unless something changes.

We're actually in agreement. Lots of my time is doing what you suggest, which is why I wish there was a time series capable version

Perhaps I've missed it but my number one request for these tools is to support time series data and analyses (or other data "structures" in general). The focus on beautiful visualization and easy graphical interfaces is great, but thinking outside the tabular box would be very helpful.

I think it's telling that someone who "only" orchestrated one of the most used pre-training and benchmark datasets in computer vision is seen as less accomplished than those who developed algorithms that require those datasets to work.

Yes, dataset creation and curation is less glamorous, but it's important work. I've worked with a few modelers who could stand to learn that first hand.

Can you explain how the actual taste component is indeed virtual?

The article mentions the hydrogels already contain the tastants, even those beyond the 5 fundamental ones, which means a rich taste+smell combo already present in physical form in the gels.

There's no virtual flavor here.

>There were nine taste-generating channels filled with flavored hydrogels. The gels were made out of agarose mixed with a bit of mineral water and specific flavor essences: sugar, salt, citric acid, cherry, milk, green tea, passion fruit, durian, and grapefruit.

Don't get me wrong, the miniturization and delivery is cool, but nothing virtual here.

The one I know of is terrifying and usually lethal. It gets in through the nose/cribriform plate https://en.wikipedia.org/wiki/Naegleria_fowleri

First, consider that the culture of the city you're in can play a role https://www.paulgraham.com/cities.html But more than that, the neighborhood or whatever natural social circle you've found yourself in is that way, but it need not be that way with some extra digging into other subgroups of people around you.

Second, consider your own behavior and whether it's not tech itself but another aspect that you're confusing with being in tech, simply because people are (annoyingly) too polite to tell you.

In "tech" I find almost as much diversity of personalities and quirks than in the general population. I haven't personally experience what you're talking about, but that doesn't mean you're wrong. Perhaps the nearest neighbors that are lowest friction to socialize with aren't the best choice?

This is all well and good, but unfortunately depends on the people pushing for the metric/system to give a shit about what the metric is supposed to improve. There are still far too many that prefer to slap 1 or 2 careless metrics on an entire team, optimize until they're promoted, then leave the company worse off.

> The best way to do that, I think, is to do away entirely with the symbolic and mathematical foundations, and to derive what Gaussians are, and all their fundamental properties from purely geometric and visual principles. That’s what we’ll do in this article.

Perhaps I have a different understanding of "symbolic". The article proceeds to use various symbolic expressions and equations. Why say this above if you're not going to follow through? Visuals are there but peppered in.

Yes and no. An alternate perspective is that the output of each neuron in an artificial neural net is analogous to an F-I curve in a real neuron (spike frequency-input DC current curve). In this way, different neurons have different slopes and intercepts in their FI curves, just as a network of ANN neurons effectively have their activation functions tweaked after applying weights.

I usually only say this to other neuroscientists who have a background in electrophysiology. The analogy isn't perfect, and is unnecessary to understand what ANNs are doing, but the analogy still stands.

There's even an artist that made this point: Cartoon Neuron

https://www.redbubble.com/shop/ap/6229477

https://x.com/Cartoon_Neuron

IME notebooks in VS Code are even worse (but improving). Jupyter lab is faster...but that depends on how fast you prefer ;)

Those comments are based on the original title introducing it as an ML Leetcode. The title is more accurate now.

So it is a leetcode equivalent ;)

In all seriousness, I think you're lucky then.

(I can only speak for my experience and those of my peers in my field, at the end of the day)

Many people who are into homeopathy have no notion of "mechanism". They don't know to ask such questions, or perhaps they don't care to.

In the worst case scenario, they accept lack of mechanism because "there are things science doesn't know", especially if their personal experience (random charlatan demonstration) is duped into belief.

Fair. More specifically what I meant was how can one avoid being subject to being pulled away from providing real value when superiors are wooed by hype. In that sense just ignoring it is not easy if you care at all about your work.

Genuine question on your preferred use-cases: Would you be comfortable with some probability of misinformation/mistakes/inaccuracies? Or would you accept the risk of some wrong information as a reasonable time vs. accuracy trade-off?

> But give it 5-10 years, or maybe less, and that will seem silly. People will be using LLMs for focused use cases and making sure they help the end users. You will be evaluated with a higher standard - of what you actually accomplish with your products.

This day cannot come fast enough in my opinion. Hype is tiring and ultimately detrimental in the long term (more disappointment than fulfilled wishes), in terms of opportunity cost of what could have been invested in instead.

How does one escape this?