Tyler Coleman

"AI fluency = critical use + applied use" A.I. can be both intimidating and misunderstood (in what one should and should not expect from it). One example is that LLMs cannot be 100% reliable and safe *by design* , which means we should know when to use them, how, and more importantly in which context. I've been trying to formalize the steps it takes from AI litteracy to AI leadership, and I stumbled on this publication that summarizes it way better than I would (with our without an LLM btw :-) ) "A Five-Step Progression to AI Fluency (For Adults and for Students)" (link in the first comment) 👇

24

2 Comments

There is very little that screams "bad writing" (or possibly "bad management notes") louder than a character doing something dumb and completely out of character because if they don't, the plot doesn't happen. Any plot that falls apart without getting goosed by a character doing something that is both stupid and something that character demonstrably would never do is a bad and weak plot. The brilliant and meticulous scientist who suddenly goes "enh" and leaves the experiment untended and bubbling overnight? The serious and hyper-competent professional soldier who leaves a gate open because "nothing's going to happen?" These are all cases where the writer has gotten themself into a hole (or been pushed into one) and the only ladder out is made out of broken character motivations. It's obvious, and it's annoying, and it's a real good way to break immersion for anyone who's been paying attention. However, I didn't just say "stupid". I said "stupid /and/ out of character", and there's a massive difference. There are actually plenty of good reasons for even smart characters to do stupid things, and they can work wonders if they're applied and presented properly. Let's start with the obvious: People don't always make good decisions under pressure. This is as true for made up people as it is for real human beings. If someone is stressed, if someone is rushed, if someone is in danger, or if someone is operating from incomplete information? You can damn well bet they're not going to always make the same call they would from a bunker a mile away with hours to discuss and research their next move. Things happen in the heat of the moment. Characters panic and react instinctively for the sake of self-preservation. This can lead to bad results, even from the best of intentions. Why? Because very, very few people are processing at their best when there's a giant ravening monster bearing down on them at forty miles an hour. There's also the possibility that the character is doing the dumb thing for reasons that are in character. The audience may know something's a bad idea, but the character might have less information than they do. They might be clinging to an idea that's not true, or they might the dumb idea is actually the right one. They might be forced into the dumb thing - pull the trigger or your loved ones get it. They might be self destructive. The list goes on and on, and any of these can save a story from "because the plot said so" /if they are set up ahead of time/. If we've been shown multiple times that a character is looking to martyr themself, then it makes sense when they do so unnecessarily. If we've been shown that a character is weak, then it's believable when that weak character takes the easy way out, even if the real cost is higher. So yes, the stupid move can be the right move for the story, but only if it flows from what the story has already told the audience. By itself, it's odd and improper. In context, it can do wonders.

71

33 Comments

I recently read an article about how Blender is funded, and it hit home, because it matched the first two comments we got the moment we announced Keyframe: "I'm not paying for a subscription." I understand it. We're in an era where the big juggernauts have people stuck in subscriptions they don't want, and most of us are carrying far too many of them. Subscription fatigue is real and it's earned. But there's another side to this that gets me excited. There are developers doing brilliant work and giving it away, or close to it. The Blender Foundation. DaVinci Resolve. Affinity, now under Canva. Free tools, top of the industry, built so people can do professional work without a professional budget. The idea is simple: if you love it, you support it, and that keeps the tools getting better for everyone. The catch is what happens when we go black and white on it and want everything free. Blender is the perfect example. Only around half a percent of the people who use it end up giving anything to the Blender Foundation. And marketing, resources and paying people all cost money, so the tool we all love ends up at a disadvantage against the companies with billions behind them. So this week's video is all about that. Not to point fingers, but to start a conversation. If you support a free tool, what made you start? If you don't, what would change your mind? I'm super curious to hear how people feel about this one. New video: https://lnkd.in/eZ7qdYsv #Blender #IndieSoftware #Animation #GameDev #SubscriptionFatigue

5

Ok GDC is fast approaching. Lets do a how to approach people for funding at GDC 101 (from a funders perspective) for those that are new to this, and some who frankly should know better. The do's: Get a meeting on the calendar someplace quiet. If at all possible. Undivided attention is the best attention. People are swamped busy all week, its likely you will end up on a vid call either before or after the conference. Accept it with grace, there is only so many hours in the day and energy to go around. Trust me when I say its not personal. Make sure you have your pitch script down. Have your elevator pitch memorized. Your first impression is key in a high pace, high volume environment like GDC. Don't smell, I know it can get sweaty running around, bring extra deodorant, do not bathe in axe body spray. BREATH MINTS. Research who you are pitching to, understand their background, understand their business culture (are you shaking hands, business card ritual, hugging), understand what their studio typically funds. Tailor your pitch script to the person and place. Have your tech charged, bring back up batteries so you can charge in-between meetings or in transit. If you are using a laptop, make sure you have a PDF backup on your phone. Video with sound always fails, optimize for video that has an easy to read visual language to get your point across. Give your self time to be on time. It always takes longer to get places, avoid back to back meetings unless people are coming to you. Even if its a short walk, take time to decompress, evaluate what went well and what bombed in the room, and make adjustments. Networking is key, face to face meetings are great. But if you are good at networking you can 2x to 3x your total meetings. Keep your LinkedIn app on the QR code, or have a business card easily available, make it super easy to lightweight swap information. Then make sure you follow up post conference with people. The Donts: Dont try to pitch people you do not have a scheduled meeting with in hallways/ corridors/ restaurants/ elevators/ in the bathroom (yes this has happened to me). This will piss off, and alienate the people you want to fund your project. Lightweight swapping of information and follow up is the goal. Just not at the urinals please. Is the sun down? Then shut it down. Look GDC is a social event for everyone, and the evening time is typically the time folks devote to their friends. Dont pitch people after dark. They are not going to remember it fondly even if they are polite enough to not stop you outright. Look if its late I will buy you a cola/coffee/ cocktail ect... but lets keep it light and social until the proper time. This goes double for anywhere there is a fucking DJ playing music (why is this still a thing?) Or when people are eating.

65

3 Comments