←
Danny2,502 posts
Danny avatar

Danny

@dazuck

Operator, founder, dad. Tech that serves people, not platforms.

◷ Joined May 2009▧ 94 Photos and videos
Kevin Roose avatar
⟳ Danny retweeted
Kevin Roose@kevinroose·Sep 25

I’m sorry, it’s a fun product, but if you hand all your data to Muse and don’t expect it to be used for the most invasive things imaginable you are Charlie Brown with the football and you deserve whatever happens to you.

Asking Meta's Muse what happens to my data:

Post attachment
Danny avatar
Danny@dazuck·Sep 24

How soon can we sign the petition to ban this in schools? @LetGrowOrg @JonHaidt

Meta announces Muse Charm, a handheld gadget allowing people to utilize an AI personal agent without their phone, computer or glasses. Available in time for the holidays.

0:11
Proton Pass avatar
⟳ Danny retweeted
Proton Pass@Proton_Pass·Sep 24

Tamagotyourdata.

JUST IN: Meta unveils Muse Charm, a Tamagotchi-style AI gadget designed to fit on a keychain, with Zuck calling it a “joyful little device.”

Post attachmentPost attachment
Danny avatar
Danny@dazuck·Sep 18

Is there literally anyone who reads this and doesn't roll their eyes at the irony and dishonesty? It actually surprises me he can write this. That there's any constituency left still willing to believe a word.

Last month I wrote about how we can build a positive and safe future for everyone: meta.com/thefutureisforevery… Every lab has the responsibility and incentive to move at the pace required to train its models safely, and the ability to take its own actions to ensure that happens. The reality is: - People won't want to use agents that are misaligned with them and that don't do what they ask, so labs have a strong natural incentive to make their models more aligned. There is a lot of debate about slowing progress on capabilities until alignment catches up. My view is that trust and alignment are quickly becoming the most important capabilities that will differentiate agents and models. Any lab that doesn't focus on alignment will fall behind. - Labs face significant liability if their models cause harm, so they have a strong incentive to prevent this as well. Meta delayed shipping Muse for several months to focus on safety and security. We didn't call for everyone else to do this before we would. We just did it as part of our day-to-day work because it was clearly the right thing for people and for us. I'm proud of the security foundations we've built. - Engaging independent evaluators and advisors is industry best practice. MSL already does this today in several areas because it helps produce better work. Other labs can just do this too. In general, it would be helpful for there to be a larger and more diverse ecosystem of evaluators. - Committing the significant majority of compute towards serving people rather than racing towards recursive self-improvement is one of the best ways to ensure we develop this technology safely. Meta has made this commitment and other labs can do this as well. I believe the key to building a positive future for everyone is maintaining the right balance of power. This is within our power to do.

Danny avatar
Danny@dazuck·Aug 1

👏👏 to whoever papered NYC with these

Post attachment
Danny avatar
Danny@dazuck·May 13

I kinda feel for employees protesting mouse-tracking by Meta. But also...you worked on behalf of Meta, the pinnacle of surveillance capitalism. What did you think, Facebook would be fine screwing over users for two decades but protect employees?

Danny avatar
Danny@dazuck·Mar 20

Companies that figure out which intelligence is theirs will ship mind-blowing software at mind-blowing speed to market it. Companies that don't may find themselves waiting in fear of each new Anthropic release. Full piece:

Danny@dazuckMar 20

Danny @dazuck Mar 20 Article Software is the new marketing. Where is the product? Recall Labs CEO Andrew recently pointed out that "software is the new marketing." As software's cost to produce rapidly falls, moats for pure software are disappearing. This is increasingly true for

Danny avatar
Danny@dazuck·Mar 20

Intelligence in the network. There's real value in coordinating across actors. @jaminball 's framing on sources of truth captures this: "The more we automate, the more important it becomes that someone has done the unglamorous work of deciding what the correct answer is." At root these are databases with software and governance on top, and the ability to extract for the software is eroding rapidly. Trust & the governance layer will provide some stickyness - especially in domains that cross organizations - but maintaining margins will be difficult.

Danny avatar
Danny@dazuck·Mar 20

Intelligence in people: the SaaS playbook is inverting. Software as a Service → Services via Software. @trailofbits open-sourced security tools that prove they understand attack surfaces better than anyone — so they get hired for the audits. @patio11 went from $100/hr to $30K/week. The SaaS products were always proof of work for the consulting. @simonw co-created Django, built Datasette, now consults on AI — millions of readers built from shipping in public. When anyone can build the tool, judgment is the product.

Danny avatar
Danny@dazuck·Mar 20

But models won't eat all, and there will be some intelligence in proprietary data. Physical-world data resists absorption — the process that created it can't be replicated by a model. Tesla and Waymo's self-driving training sets. Recursion running 2.2M wet-lab samples per week. The bar for hard to reproduce data is getting way higher, and many will fail to find footing here. But some exists.

Danny avatar
Danny@dazuck·Mar 20

A lot of the intelligence will be in the models. When Anthropic launched a Meta ads connector, @irabukht (refreshingly honestly) said Ryze AI's close rate collapsed from 70→20% in weeks. They aren't alone. Jasper's revenue roughly halved. Chegg visits dropped 89%. Textual, derivable data is getting absorbed fast.