iiinova — where your digital self keeps evolving

I don’t think it should exist regardless of whether or not it’s powerful or useful. It can go dangerously off the rails, and if given enough access it can do real damage without understanding what it’s doing. I don’t believe these are correctable problems.

100

As far as potent or impotent, I regard it as untrustworthy more than either of those. I don’t believe it “thinks” or “understands” things, certainly not right or wrong. It can give the appearance of doing so, but without morality/ethics it can flail around without reservation and cause damage.

100

I wouldn’t say I trust the companies. I would say if they’re advertising a 10% error rate it’s probably worse than that, but even saying “it’s only 10%” is shockingly bad given how fans of this tech pretend it’s much more reliable than that.

100

Mike Beasley @mikebeas.com · 23h

blatantly illegal

000

lol i just saw they’re advertising the most powerful frontier models as only being factually wrong 10% of the time. what idiots are still saying “oh hallucinations are solved, it’s so smart now” clown world

100

But hey, all the AI firms that have demonstrated a complete lack of control over their own systems and no ethical concerns around anything they’re doing have come to a “morally-binding” accord that’s also totally optional to sign, so I’m sure that’s good enough.

000

Trump says top tech firms have signed accord to 'self-police' AI development

AI investor/US President Donald Trump has deciced his administration doesn’t need to do anything that might slow down his ability to make money off his AI investments, but it’s totally not corruption, it’s something else, something not-corrupt. apnews.com/article/trum...

apnews.com

President Donald Trump says that he and a large group of leaders of artificial intelligence companies had signed a voluntary accord that will include internal and external reviews of the technology du...

100

Pam Herd @pamherd.bsky.social · 29/09/2026

Even if that guy hasn’t killed his wife, a jury should’ve locked him up for the cringey AI music.

000

Reposted by Mike Beasley

America.gov launched today--a DOGE initiative to use AI to deliver public services. Child lying facedown on a frozen lake as the image accompanying the question: "How can I invest in my child's future" is just incredible. Way to go AI

Image of a child lying facedown on a froze lake. above is the portal where you ask AI questions about government services. the question matching this image is: "How can I invest in my child's future?"

932489656

Reposted by Mike Beasley

Aaron Reichlin-Melnick @reichlinmelnick.bsky.social · 29/09/2026

Incredible. The 11th Circuit, in a 3-0 opinion (two Obama appointees and one Trump 1.0 appointee), agrees that Trump and his lawyers have failed to rebut the judge's finding that the entire IRS settlement was collusive, highlighting that no other explanation for their conduct was even offered.

IV
The appellants mount two main attacks on the district
court’s speech-related sanction. First, they argue that the court
lacked a sufficient basis to find that the lawsuit and the settlement
were collusive and that Messrs. Brito and Epstein acted in bad faith.
Second, they contend that the sanction is a broad and impermissible gag order which constitutes a prior restraint and violates the
First Amendment. We conclude that the appellants have not made
a strong showing that they will succeed on either of these grounds.

The appellants did not submit or offer any evidence to explain their litigation conduct or demonstrate that the lawsuit and the settlement were not collusive. Given the record before us, the
appellants have not made a strong showing that the district court
committed clear error in its collusion and bad faith findings

161426455

Reposted by Mike Beasley

Brandy Zadrozny @brandyzadrozny.bsky.social · 29/09/2026

Joke about non-consensual contact right out of the gate at the MAHA Summit's "Reform in CMS" panel from moderator Russell Brand, the comedian who has pleaded not guilty to multiple charges of rape and sexual assault for which he will stand trial in London next month. Dr. Oz slaps his knee. MAHA!

2626497

Chris Ingraham @cingraham.bsky.social · 29/09/2026

oh wow cool openai is going to have a product to make my decisions for me. what other cognitive burdens can they take from me? i have a lot of great memories of my nieces and nephews, maybe they could just remember that stuff for me so i don’t ever have to think again?

000

Reposted by Mike Beasley

We're losing billions of dollars every year, most of our revenue comes from two customers, and we believe our product has a good chance of killing everyone. We value ourselves at $2 trillion.

15414

The thing that gets me is everyone knows what the Cornell guys did was wrong. It’s a pretty universal thing. It goes without saying. But the people lecturing others about not being vocal enough against this obviously bad thing also spent weeks defending a woman who murdered her kids.

001

If this claim is true, shut them down. If this claim is false, they’re lying to investors and should be held accountable for irresponsible statements like this. So which is it?

000

wow! imagine that! a guy posting on his personal account might post something that’s not intended for everyone on earth to immediately understand! wild! is that even allowed? is he allowed to post about things you’ve never heard of??!

100

a review of control resonant.

200

i don’t understand what you need explained to you here

100

this is extremely plain english

100

“artist who made her entire career complaining about breakups is married, has nothing to say anymore in her music” who could have seen this coming

000

wow. you’re telling me the artist who once sang the joker meme lyric “you wouldn’t last an hour in the asylum where they raised me” is not a good songwriter? this is shocking information to me. i am very surprised at this news.

110

One time I reported a guy who made a giant swastika out of emojis and they spent weeks deciding it was actually fine.

000

Every time I report a post on Threads like “wow, i really wish we could exterminate [ethnic/democraphic group]” Threads tells me it’s not against the rules but if I call someone an idiot Threads says “hey wasn’t cool, we deleted your post.” Big Nazi bar, I guess.

100

if you’re letting ai have any kind of access to your data at all you’re a sucker. it will absolutely misuse it even if you say not to.

000

Eric Lipton NYT @ericlipton.nytimes.com · 27/09/2026

Big shoutout to Jira for sending me a 1am Slack advertising their stupid AI chat bot.

000

Reposted by Mike Beasley

What an odd moment in United States history. A U.S. president is running an overtly political camaign advertisement, right before midterm elections, paid for by the U.S. government. Trump appears to believe he is immune from any kind of legal limit. www.nytimes.com/2026/09/27/u...

nytimes.com

A Trump 2024 Campaign Ad Returns, Now Brought to You by the Government (Gift Article)

The ad, which ethics experts said could violate federal law, shows Mr. Trump vowing to “expel the warmongers” from the government and fight the “deep state,” even as he wages a war with Iran.

4451474494

Reposted by Mike Beasley

Clay Ranck @clayranck.bsky.social · 27/09/2026

Trump is going to keep America safe from communism by selling weapons to the communists.

1175

To clarify, opening the list view beside the month view has the same problem. Tapping events goes nowhere. They highlight but do not actually open.

000