Other platforms too, but I'm on lemmy. I'm mainly talking about LLMs in this post

First, let me acknowledge that AI is not perfect, it has limitations e.g

• tendency to hallucinate responses instead of refusing/saying it doesn't know
• different models/models sizes with varying capabilities
• lack of knowledge of recent topics without explicitly searching it
• tendency to be patternistic/repetitive
• inability to hold on to too much context at a time etc.

The following are also true:

• People often overhype LLMs without understanding their limitations
• Many of those people are those with money
• The term "AI" has been used to label everything under the sun that contains an algorithm of some sort
• Banana poopy banana (just to make sure ppl are reading this)
• There have been a number companies that overpromised for AI, and often were using humans as a "temporary" solution until they figured out the AI, which they never did (hence the gag, "AI" stands for "An Indian")

But I really don't think they're nearly as bad as most lemmy users make them out to be. I was going to respond to all the takes but there's so many I'll just make some general points

• SOTA (State of the Art) models match or beat most humans besides experts in most fields that are measurable
• I personally find AI is better than me in most fields except ones I know well. So maybe it's only 80-90% there, but it's there in like every single field whereas I am in like 1-2
• LLMs can also do all this in like 100 languages. You and I can do it in like... 1, with limited performance in a couple others
• Companies often use smaller/cheaper models in various products (e.g google search), which are understandably much worse. People often then use these to think all AI sucks
• LLMs aren't just memorizing their training data. They can reason, as recent reasoning models more clearly show. Also, we now have near frontier models that are like 32B, or 21B GB in size. You cannot fit the entire internet in 21GB. There is clearly higher level synthesizing going on
• People often tend to seize on superficial questions like the strawberry question (which is essentially an LLM blind spot) to claim LLM's are dumb.
• In the past few years, researchers have had to come up with countless newer harder benchmarks because LLMs kept blowing through previous ones (partial list here: https://r0bk.github.io/killedbyllm/)
• People and AI are often not compared fairly, for isntance with code, people usually compare a human with feedback from a compiler, working iteratively and debugging for hours to LLMs doing it in one go, no feedback, beyond maybe a couple of back and forths in a chat

---

Also I did say willfully ignorant. This is because you can go and try most models for yourself right now. There are also endless benchmarks constantly being published showing how well they are doing. Benchmarks aren't perfect and are increasingly being gamed, but they are still decent.

> Rep. Ro Khanna (D-Calif.) put Republicans on the spot with the introduction of his Drain the Swamp Act, a bill aimed at banning White House officials from accepting gifts from lobbyists and preventing them from becoming lobbyists.

> The bill directly challenges Trump to uphold his long-standing campaign promise to "drain the swamp" by eliminating government corruption.

> President Trump campaigned around the country to 'drain the swamp', yet one of the first things he did was reverse President Biden's executive order that banned White House officials from accepting gifts from lobbyists," Khanna said on the House floor. "I believe that this bill will have support, not just from progressives, not just from independents, but from the MAGA movement."

> Khanna's move forces Trump-aligned Republicans to either support stricter ethics reforms—aligning with Trump's past rhetoric—or reject the bill, which could be seen as backtracking on promises to clean up Washington.

> Last month, Sen. Elizabeth Warren (D-Mass.) accused Trump of breaking his promise to "drain the swamp" during his first term in a letter urging him to address "key corruption risks," a likely reference to Elon Musk, who holds a government role while maintaining extensive private business interests.

> "The American people have seen that, all too often, government officials use their positions to benefit their own pocketbooks," Warren wrote. "Even the appearance of such corruption is enough to damage Americans' trust in government."

> Khanna's bill is the latest effort from Democrats to test whether Trump and his allies are willing to follow through on anti-corruption rhetoric—or if "draining the swamp" was just a campaign slogan.

(I haven't submitted an official rfc yet, want to see what people think)

This is inspired by Ruqqus, a now defunct Reddit alternative.

The idea is simple:

1. There is a "global" or "default" community with no topic or extra rules, moderated only by admins
2. Community moderators, when they feel a post is inappropriate for their community can "kick" a post to the global community

The reasoning is as follows: a good amount, probably the majority of posts that are removed by mods, are not removed because they are inappropriate for the site as a whole, but because they are inappropriate for that specific community (off-topic, banned site, low effort, etc.). But currently the only option they have to deal with this is a full blown removal, which is quite frustrating for the poster.

This proposal would allow mods to keep curated communities without needing to do unnecessary removals.

---

As a bonus, this would create a default community where people can post when they're not sure where to post something. Posts can be later be crossposted into more specific communities.

authentic content my hat

I was unable to upload even the shortest video because it was too long for my instance. Therefore, please enjoy the following:

1. Partial visualization of test data. I cut this short because it took 40 seconds to do just a few (out of 81) paths: https://youtube.com/shorts/7UvzgSsMQNA
2. Partial visualization of full data. I cut this short because I didn't want to wait 40 minutes. It's sped up 2x by making it 60fps (each step is approximately one frame) https://youtu.be/cv9qSdrV2Z4
3. Full visualization, but it only shows the end paths, not individual steps: https://youtube.com/shorts/ozQ77ikI7JI

Unfortunately youtube is forcing my videos to be shorts due to aspect ratio and length, I don't know if I can force them to a regular video

Full input link (it's kinda cool ngl): https://youtu.be/Jw3CXcaHZ0Q

I thought about it for 15 mins, but couldn't think of any mathematical tricks. I thought of lots of minor tricks, like comparing the number to the result and not adding any more multiplications if it's over, things that would cut 10%-20% here and there, but nothing which fundamentally changes big O running time.

For reference, here's my solution for part 2 in smalltalk. I just generated every possible permutation and tested it. Part 1 is similar, mainly I just used bit magic to avoid generating permutations.

(even if you haven't used it, smalltalk is fairly readable, everything is left to right, except in parens)

```smalltalk day7p2: in | input |

input := in lines collect: [ :l | (l splitOn: '\:|\s' asRegex) reject: #isEmpty thenCollect: #asInteger ].

^ (input select: [ :line | (#(1 2 3) permutationsWithRepetitionsOfSize: line size - 2) anySatisfy: [ :num | (self d7addmulcat: line ops: num) = (line at: 1)] ]) sum: #first. smalltalk d7addmulcat: nums ops: ops | final |

final := nums at: 2. ops withIndexDo: [ :op :i | op = 1 ifTrue: [ final := final * (nums at: i + 2) ]. op = 2 ifTrue: [ final := final + (nums at: i + 2) ]. op = 3 ifTrue: [ final := (final asString, (nums at: i+2) asString) asInteger ] ].

^ final ```

Full input version here: https://youtu.be/itc7_HZmewQ

This poor guard must the fittest person in 1518. Man's doing a marathon every day

I had to stop for finals, so I'm way behind