@ovaska

Sharing the adventures of an AI-First founder and product leader. Two exits. Former photojournalist. Advocate for truth.

 Brooklyn, NY
ovaska.com
Joined December 2008
  • Tweets2.7K
  • Following653
  • Followers323
130Media
In the absence of doubt there is tyranny.
7
If rollercoasters were universally and infallibly safe they would also be much less fun.
1
29
AI has wrought a completely unsustainable build to strategy ratio.
1
1
31
When coding was expensive, planning and strategy was essential to keep costs down. Now planning and strategy largely happens after delivery. This feels extremely fast but it’s a mirage.
16
Selling the product vision starts post-acquisition.
1
1
30
Because the M&A process involves so much selling, it's easy to think that everyone is sold when the deal closes. (They bought it, after all!) In my experience, this isn't the case.
21
The most is human thought, relationships (network as you say), and data.
Nikita Bier
@nikitabierOct 12
Technical moats are mostly over at this point. If you’re building something that is meant to be durable, the only moats now are: • Proprietary data • Network effects • A physical component
1
36
Give professional feedback in real time. The longer you wait the more it’s about you and not them.
2
3
170
I’m always surprised by how many otherwise good leaders don’t hone this most fundamental part of leading.
21
Writing a weekly series on how I made my independent feature. It’s advice and tips on how you can make a movie and create independently, BTS if you’re interested in film or just curious how it works. Please repost! This is community!
2
29
56
1,884
A product company death spiral goes like this: The product gets some traction. Company demands growth. Product leader goes broad thinking more product, more better. Costs balloon, market is confused. Company is sold for scrap.
1
2
56
Go deep. Double down on the thing that seems to work. If you pivot make it a hard turn and go all-in there. Don’t go broad until you dominate the first thing.
34
Products have dozens of special interest groups. It’s essential to find the “anchor” group that defines where your product’s responsibility is focused. Approach to every other group can be half-baked and you can still win.
1
23
Lying to yourself changes how you remember the past. You essentially live a fantasy.
1
1
38
This is a powerful accelerator in good or bad directions. In teams it’s important to state firmly when something is not true or ask for clear evidence.
16
Important work
Tom Maiaroto
@tmaiarotoSep 26
Early thoughts and work on agent memory poisoning protection... github.com/hackafterdark/pho…

phosphor/docs/memory/POISONING.md at main · hackafterdark/phosphor

Contribute to hackafterdark/phosphor development by creating an account on GitHub.

github.com
1
49
A single well organized engineer can now easily out deliver several product folks working on strategy and stories. Some engineers use this to have more free time. Some are using agents to do the job of product faster. Some are skipping product entirely and shipping on speculation. Product folks, for their part, have (AFAICT) responded by using AI to generate slop stories or (if the long term audacious goal is clear) pushing for more faster and getting fast and loose with stories. None of this is sustainable. This is trying to fit SCRUM into something completely different. The entire SDLC is different. The effort and importance of every function has changed. It’s time to go back to 1998 before agile and scrum and… with a clean mind… consider how the best software should be made.
38
Product success now will come from team coordination to better understand the need and context so that what is built is better, faster and cheaper than competing teams/companies. AI-first team members need to bring their strategic life experience and not traditional roles like product, BA, QA, engineering.
1
1
42
The questions that are now important for individuals and teams are: what are the new roles of Product, Eng, QA, etc? How does code factor into the _business_ when you can code basically anything as fast as you can imagine it for a nominal cost?
25
I lived through the WYSIWYG revolution where hand coding HTML became artisanal. Die-hards held out claiming quality and reliability and compatibility by coding by hand. Same thing is happening at a massive scale with coding in general.
1
1
2
68
Never stop learning. So many of my conversations about the impact of AI are frustratingly lacking in curiosity about what’s happening and about to happen.
DHH
@dhhSep 24
"In a time of drastic change it is the learners who inherit the future. The learned usually find themselves equipped to live in a world that no longer exists." — Eric Hoffer, 1973
1
31
Load more