We’re thrilled to announce Skip 1.0!
Skip brings Swift app development to Android. Share Swift business logic, or write entire cross-platform SwiftUI apps, with native performance and native UI on each platform
Now free for Indies!
https://t.co/UA3pNWHbul
I was wondering about the colors of the #aurora. I saw this unattributed image floating around and a post from @Astropartigirl claiming it has to do with the concentration of oxygen. But that didn't make sense to me. A short thread 1/N
[attribution for in image description]
Apple's new "Crush" ad (let's call it "2024") is a visual & metaphorical bookend to the 1984 ad.
1984: Monochome, conformist, industrial world exploded by colourful, vibrant human
2024: Colourful, vibrant humanity is crushed by monochrome, conformist industrial press
Eclipse Reminder - Earth is the only planet that we know of that has a perfect solar eclipse, since our moon happens to be about the exact same size as our sun when seen from the surface.
As far as we know this is a coincidence, and about a one in a trillion bit of luck.
# On the "hallucination problem"
I always struggle a bit with I'm asked about the "hallucination problem" in LLMs. Because, in some sense, hallucination is all LLMs do. They are dream machines.
We direct their dreams with prompts. The prompts start the dream, and based on the LLM's hazy recollection of its training documents, most of the time the result goes someplace useful.
It's only when the dreams go into deemed factually incorrect territory that we label it a "hallucination". It looks like a bug, but it's just the LLM doing what it always does.
At the other end of the extreme consider a search engine. It takes the prompt and just returns one of the most similar "training documents" it has in its database, verbatim. You could say that this search engine has a "creativity problem" - it will never respond with something new. An LLM is 100% dreaming and has the hallucination problem. A search engine is 0% dreaming and has the creativity problem.
All that said, I realize that what people *actually* mean is they don't want an LLM Assistant (a product like ChatGPT etc.) to hallucinate. An LLM Assistant is a lot more complex system than just the LLM itself, even if one is at the heart of it. There are many ways to mitigate hallcuinations in these systems - using Retrieval Augmented Generation (RAG) to more strongly anchor the dreams in real data through in-context learning is maybe the most common one. Disagreements between multiple samples, reflection, verification chains. Decoding uncertainty from activations. Tool use. All an active and very interesting areas of research.
TLDR I know I'm being super pedantic but the LLM has no "hallucination problem". Hallucination is not a bug, it is LLM's greatest feature. The LLM Assistant has a hallucination problem, and we should fix it.
</rant> Okay I feel much better now :)
Programmers who can add functionality, without making the codebase unnecessarily complicated, are *incredibly* valuable. If someone has properly absorbed this aesthetic and knows how to do it, I would hire that person on the spot regardless of perceived lack of other skillsets.
“You can come with your kippahs and you'll face nothing but love from the people here.”
Jewish protester, Maudi, who attended the pro-Palestine march in London on Saturday encouraged fellow Jews to show their support and attend demonstrations while wearing their kippahs
Urgent appeal to UN-Secretary General Mr. Guterres:
Break the siege of Gaza NOW!
Time is critical: Establish an immediate airlift with lifesaving supplies to the besieged people of Gaza.
@antonioguterres@UN#BreakTheSiege
Max Planck was a musical prodigy who could play the piano, organ, and cello. He even composed his own songs and operas. He once said,
“The main source of all the greatest achievements in natural science, I am convinced, lies in the divine gift of musicality.”
The Agile Manifesto says we "build projects around *motivated* individuals." ❤️
Want to know the 2 main reasons why Scrum and Agile transitions fail: 🤷♀️
1. The individuals on the team are not motivated 😞
2. The individuals want to work in silos, not as a team 👯
The bigger problem?
Many in the Agile community can't come to grips with the fact that some people are unmotivated, nor can they come to grips with what to do with those people. 🤦♀️
I once suggested on LinkedIn that destructive, lazy, unmotivated and deleterious employees should be let go, and the blowback was amazing. 💣
"You should do everything humanly possible before you let an employee go" was a common response.
Of course, when I asked if they'd share their salary to keep the unmotivated employee motivated, the virtual signalling of 'doing everything' quickly came to an end. 💰 💰 💰
So what happens when you keep unmotivated and deleterious team members around?
What happens when you keep people on the team who are lazy and don't do their work?
Here's the reality: you have to manage them. 🤦♀️
✅ You have to assign people who can't self-manage JIRA tickets.
✅ You have to track the work of unmotivated people.
✅ You have to give people who can't self-manage assignments.
And you know what else?
All of this ends up spilling over to the rest of the team.
In the end, everyone gets managed. No organization hires a person to manage just one individual on a team. Everyone gets managed once a manager comes on the scene.
If you want to be Agile, and you want Scrum to work, respect what the Agile Manifesto and the Scrum Guide says.
Developers are expected to be motivated and self-managing.
If you think Scrum will work, or an Agile transition will take hold with a group of unmotivated individuals who are unwilling to manage themselves, you are setting yourself up for some significant disappointment.
👉This new paper is far more important than the apparent collapse of Cruise, but not unrelated. The unsolved challenge of distribution shift is *central* to why reliable driverless cars (& eg reliable agents) are still years away.
👉 The new paper’s conclusions are virtually identical to what I said in 1998 (below) and 2001. For a quarter century I have been confronted with aggression and denialism for saying it; that does not speak well for the culture of ML.
👉 @random_walker and @benjaminjriley are right that this new result is a big deal.