Jason Wolf (@JasonGuyWolf) header
Jason Wolf (@JasonGuyWolf) profile picture

Pro nerd. Interested in #product, MCPs, automation, RAG/LLMs, design, making stuff & learning new things. 🤖 Product @mcdean, tweets my own.

Washington, DC
Joined October 2009
Photo by Jason Wolf (@JasonGuyWolf)Photo by Jason Wolf (@JasonGuyWolf)Photo by Jason Wolf (@JasonGuyWolf)Photo by Jason Wolf (@JasonGuyWolf)Photo by Jason Wolf (@JasonGuyWolf)Photo by Jason Wolf (@JasonGuyWolf)Photo by Jason Wolf (@JasonGuyWolf)Photo by Jason Wolf (@JasonGuyWolf)
Figure it out. Make it work. #streetsofDC #mvp
2
Jason Wolf retweeted
As usual, Anthropic just published another banger. This one is on context engineering. Great section on how it is different from prompt engineering. A must-read for AI devs.
60
434
27
3,927
368,288
This is interesting as a first large diffusion-based LLM. Most of the LLMs you've been seeing are ~clones as far as the core modeling approach goes. They're all trained "autoregressively", i.e. predicting tokens from left to right. Diffusion is different - it doesn't go left to right, but all at once. You start with noise and gradually denoise into a token stream. Most of the image / video generation AI tools actually work this way and use Diffusion, not Autoregression. It's only text (and sometimes audio!) that have resisted. So it's been a bit of a mystery to me and many others why, for some reason, text prefers Autoregression, but images/videos prefer Diffusion. This turns out to be a fairly deep rabbit hole that has to do with the distribution of information and noise and our own perception of them, in these domains. If you look close enough, a lot of interesting connections emerge between the two as well. All that to say that this model has the potential to be different, and possibly showcase new, unique psychology, or new strengths and weaknesses. I encourage people to try it out!
We are excited to introduce Mercury, the first commercial-grade diffusion large language model (dLLM)! dLLMs push the frontier of intelligence and speed with parallel, coarse-to-fine text generation.
371
1,490
185
11,322
947,020
Jason Wolf retweeted
I made a trailer for the future of humanity (e/acc)
299
993
327
5,910
1,318,957
If any of the laid-off Bungie devs are making a new game dev project... If you're potentially looking to become an Indie Dev to get work going as a new studio... Talk to me. Pretty sure my viewers and I would be interested in coverage of your game.
74
441
10
6,408
337,046
Jason Wolf retweeted
Replying to @GriffinWB @Bungie
We ate the shit together on all too many occasions. You rose above and made sure shit could still happen. You created new workflows and processes to improve Social. You dealt with the headaches and made sure to stand up against toxicity and harassment. Much love, my man. 🫡
6
31
1
2,831
95,283
This is too much, I think I love it ❤️🤡🤠🔥✨ - youtu.be/dseLM0ZpS5Q?si=LgVH… @randwerk #abriss
1
1
98
🪦RIP GPUs 🥵
2
41
ChatGPT has now a big problem. Google just updated its free competitor, Bard. Here are 8 things impossible on ChatGPT but that Bard can do (for free):
1,516
11,815
1,590
59,572
12,808,668
Louder for the people in the back! The amount of times I had to follow these two companies into a company to clean up their mess when they tried to introduce #prodmgmt… and charge an exorbitant fee for nothing.
Consultancies, the economist argues, know less than they claim and cost more than they seem to on.ft.com/3jW7IOA
7
44
10
319
94,892
At McKinsey, we charged $400k+ per presentation. Here’s the simple 7-step framework we used (steal it for free)🧵:
698
4,239
276
20,871
4,116,247
Mckinsey charges $10,000 per day for a consultant at my level.....yeah👀 No its not for their problem-solving prowess but for phenomenal people skills. Here is the most powerful people skill I learned there:👇🧵
311
1,281
87
7,585
Learned a lot with @lissijean and her new book ‘Escaping the Build Trap,’ - got some 🔥 tips on product strategy for @BBGIndustry 😎 Thank you for stopping by Melissa! (Book plug: amazon.com/Escaping-Build-Tr…)
2
5
A “razor” is a rule of thumb that simplifies decision making. The most powerful razors I’ve found:
1,216
17,182
2,106
83,102
10 years ago, you had to learn PM from books and mentors. Now, podcasts drop knowledge on the regular. You’re falling behind if you don’t listen. 12 podcasts you can’t miss: 1. @lissijean
18
69
12
497
Big difference between: "When will you be done?" and "What can I do to help you?"
11 Promises from a Manager: a 🧵 1. We’ll have a weekly 1:1. I’ll never cancel this meeting, but you can cancel it whenever you like. It’s your time.
572
6,015
1,922
48,532