Corence

Google search is worse because of recent AI tech flooding the internet with misinformation and low quality articles.

https://gemini.google.com/share/3717c8505d6b

Two of the three strip titles are hallucinated and two of the three strips are bad examples. Haley is mute in strip 403 and does nothing. Strip 578 is the start of the arc that shows the behavior Gemini is talking about, but has things going wrong so it's not a good example either.

Claude picks a good strip but also hallucinates the strip title: https://claude.ai/share/56be379d-c3da-443e-b60f-2d33c374eba8

I think LLMs are better at finding the most helpful sources now, but that's more a testament to how much the front page of web search has lost to low value LLM content.

I've used it for live service video games, it's pretty good at summarizing major changes to a game since you've played it last. With regular web search you'd have to go to every major patch and a lot of games don't even have good patch notes / it's all stuck in content creator videos.

Though I still prefer Claude for this since it's better at citing sources.

Unironically you can put "what are good demos for agentic workflows at Google I/O that would be received well by the general public" into Gemini's AI Mode and get better suggestions for use cases than what they're showing.

Note the scoring function is significantly different for ARC-AGI-3. It isn't the percentage of tests passed like previous versions, it's the square of the efficiency ratio -- how many steps the model needed vs the second best human.

So if a model can solve every question but takes 10x as many steps as the second best human it will get a score of 1%.

It is a fair question. I'd expect the numbers are all real. Competitors are going to rerun the benchmark with these models to see how the model is responding and succeeding on the tasks and use that information to figure out how to improve their own models. If the benchmark numbers aren't real their competitors will call out that it's not reproducible.

However it's possible that consumers without a sufficiently tiered plan aren't getting optimal performance, or that the benchmark is overfit and the results won't generalize well to the real tasks you're trying to do.

Focusing on the AI content generation aspect of this is disappointing to me, since those tools are fuzzy at best and even if the post was AI generated, it isn't necessarily a red flag since the user could be trying to disguise their writing style.

There are plenty of other signs this story is likely fake. The author claiming to be posting from a library on New Years' Day (most government buildings are closed) and was responding over 10 hours on the account. He's using a throwaway and a "burner laptop" at the library, but he also says he put his two weeks' notice in yesterday (also odd that this is on New Years' Eve) which would make identifying him trivial.

Fake stories get to the front page of Reddit every day, I wish journalists were pointing out the actual signs not to trust something to act as a better example.

FYI: Card 8's transcription is different than the image. In the image 5, 8, 12 is a Set but the transcription says Card 8 only has 2 symbols which removes that Set.

See: "How do we know it’s training data?" from the posted link.

Tesla's example is most alarming to me since people who seemingly have no business purpose to access highly sensitive data have access to it. Culturally people feel safe sharing this data on internal chats which means they don't think coworkers will report the data access violations and since the content is spreading "like wildfire" there's a significant number of people at the company who are abusing data access as opposed to an individual abusing their elevated access.

Offworld Trading Company is a real-time economic strategy game. It does have a fair amount of things to manage, but having no units means the APM is far lower than a traditional RTS.

O, Theta, and Omega are independent from best, worst, and average case. You can mathematically establish upper, lower, or exact bounds on any of the average, best, or worst cases.

It is perfectly valid to say "the best case is O(n)" which means the best case scales no worse than linearly. The "no worse" here is not describing the best case, but rather the strictness of your bound. You could say a sort algorithm is O(n!) since, yes, it does scale no worse than n! but it's not particularly helpful information.

Big O notation is used imprecisely frequently (see also people saying a dataset is O(millions) to describe scale).

Fortunately I have a rock that says "the next example will make the same point as the example I just read" so I saved a bunch of time.

Yeah, even things like gather.town people have said worked well for things like happy hours for teams working remotely. Having some virtual equivalent of physical space helps people socialize more naturally.

My chat with agent button was kind enough to work when I was canceling my crossword subscription. When they asked why I said that I suspected their unsubscribing process to be illegal under California law and their next message was that they had unsubscribed me.

If a rule-based system can get as good of results as a deep neural net, why is the deep neural net "AI" but the rule-based system is "dumb and hard-coded"?

AI is not a precise term. If you can make a product that feels intelligent to the user, why does the implementation matter?

Yeah, the presence of piece drops means the average piece power has to be lower in Shogi. If you play Crazyhouse you see how quick the game is if you have drops in a game with as powerful pieces as Chess.

I think both games have their interesting parts and it's fun to play both.

At the end of the game, yes, but if players don't try to bankrupt each other it probably won't happen.

Early in the game the expected value of a trip around the board is positive because of low rents of undeveloped properties and the net gain from Chance and Community Chest, plus passing Go. Cash flow into the game is positive as a whole.

Later in the game, when property is developed you're going to be paying bigger rents. This is going to cause players to mortgage or sell houses to pay the big rents (also their own liquid cash supply is likely lower from developing their own properties), each of which is a money sink since you don't get full value for these. This causes money to exit the game, eventually leading toward bankruptcy.

However, this requires that players actually get to develop their properties. If the game hits a lock due to no one trading and no one getting a natural monopoly, the game can go pretty much indefinitely.

The discovery was awful too. They had a similar exclusivity deal for some Dota 2 tournaments, and I could not find the stream from the Facebook home page while it was live.

I've heard this was a problem since the spice level was optimized in an office where they make the cookie dough the day of service, but deployed in an office where they make the dough the night before and cook on demand. The spices in the dough longer made the flavor stronger than it was optimized for.

A good example of training/serving skew.

That was true for their original 1v1 bot, but in the latest blog post they mention bots can learn it on their own if left to train longer.

Second, we can bring what is known as differentiated education — based on the idea that students master skills in different ways and at different speeds — to every student in the country. A 2013 study by the National Institutes of Health found that nearly 40 percent of medical students held a strong preference for one mode of learning: Some were listeners; others were visual learners; still others learned best by doing.

This study was a 100 person survey with no experimental evidence to back up people's beliefs. I've been working with educators for a while, and they've said that "learning styles" are just a myth with no solid evidence. All this survey says is that people think they learn best via one method, but people are notoriously bad at knowing what is an effective way of learning. In surveys people overvalue the "intensity" of a learning experience [1], which makes them think things like a 3-week bootcamp are more effective, when really spaced repetition is a better use of time.

At least in the US, there are many easier ways we could improve our education methods before resorting to ML-driven customization.

[1] http://www.pnas.org/content/114/37/9854.full

My understanding is that there are two parts to this suit: patent infringement and trade secrets. You can't patent a trade secret (then it wouldn't be secret), so these are two independent things.

Alsup was suggesting Waymo to drop the patent portion of their suit and focus solely on the trade secret portion of their case. The trade secrets were pared down as mentioned in that snippet, but that's in addition to dropping 3 of the 4 patent infringement claims.

I disagree with this. In my experience, improving your dataset quality (either via gathering more examples, getting more reliable labels, or doing better normalization) has yielded bigger gains than trying to use the latest and greatest model architecture for the problem. Linear and tree models are still widely used in applied problems.

That post doesn't really make sense. Of course if you can get two Monopolies without any other player having a threat you will win. Building shortages are rarely used in serious play.

Minus-Celsius is a former tournament Monopoly player and his posts on this thread summarize it better than I could: https://www.reddit.com/r/boardgames/comments/42wze4/how_to_w...

I don't really understand the motivation behind chat UIs. Dan Grover says it better than I could here: http://dangrover.com/blog/2016/04/20/bots-wont-replace-apps....

The article says they're restricting the APM for the agent to try to reduce the impact of this.

Yes, when I began studying CS I got hit with this pressure really hard. It definitely gave me the feeling that I didn't belong in the field. It is like a cousin to Impostor Syndrome. Instead of feeling like I'm incompetent (though that still happens sometimes), I feel like I shouldn't be in the field since I don't care enough about it to work on it more than 40-45 hours a week.

Three years into my job, I'm mostly over it. I still have to deal with coworkers laughing that I use Windows on my personal computer, but at least I can eat the cake I made this weekend.