Research & design @NewsLitProject. Resources: https://xitter.cf/t.co/hW0VhvQ8bH. I'm mostly here these days: https://xitter.cf/t.co/URMOp567un

Chicago
Joined June 2009
Followers: If you're also flying the blue skies, you can find me here: bsky.app/profile/peteradams.…
1
3
13
707
Not a bot. I’m a real human. I’m a chemist. And apparently defending science is controversial on X. Vaccines save lives. Chemistry saves lives. Evidence matters more than opinions. Yet scientists are often drowned out by people who have never set foot in a laboratory. I think it’s time more scientists started speaking up. Chemists. Biologists. Physicists. Doctors. Researchers. Engineers. We need your voices outside the laboratory. So here’s a little experiment. If you work in science, introduce yourself. Tell us what you do. 👋 Let’s remind X how many real scientists are out there. Bonus points if you have a dog too! 🐶
327
716
18
4,037
32,245
An AI-enhanced version of a grainy image that appears to show several men being arrested in relation to the ongoing incident at RAF Fairford is currently being shared on multiple platforms. A SynthID check reveals that the enhanced image was generated with OpenAI (first image). In the grainy version, which seems to be genuine, the faces of the men being arrested, along with a few other details, are not clearly visible (second image). Enhancing grainy images and videos with generative AI to identify key missing details - such as faces, writings, locations, logos, etc - often produces inaccurate and unreliable results. That's because AI models have no reference as to what the missing details in a real but grainy image or video look like, and therefore, they often generate those details based on an incorrect guess rather than factual data.
22
122
14
440
56,668
This will be a long thread. Please stick with it. The Miami Herald, one of the premier legacy newspapers in America, laid off or wrote off more than 30 positions today – well over one-third of its editorial staff. The cuts include crucial accountability beats. (1)
154
1,130
94
3,056
449,851
The Herald reporter who broke that story is no longer here. Her replacement in the Capital is no longer here. The job likely will be eliminated. Of what significance is government in the sunshine when there are no watchdogs to look in all those otherwise dark places? (7)
4
94
2
609
41,470
Some of the content generated by the latest AI models are frighteningly realisitc, to the point that it's becoming increasingly difficult to spot any obvious errors in them with naked eyes, even for those with eyes well-trained on AI content. Going forward, digital watermarking and identification technology seems to be one of the only practical solutions to AI-generated misinformation. It works by embedding watermarks in any content generated by an AI model at the time of creation, which will then allow a verification tool, preferably developed by the owner(s) of that same AI model, to detect those embedded watermarks.
14
84
3
283
55,865
People say X is defending free speech. Look at what actually gets amplified. Not nuance. Not expertise. Not evidence. Outrage. Conspiracy theories. Propaganda. X isn't a free speech platform. It's a machine that turns anger into engagement and engagement into profit.
100
155
16
755
12,622
Science tries to prove itself wrong. Pseudoscience tries to prove itself right. One follows the evidence. The other follows the conclusion.
38
68
3
399
6,094
Science didn’t lose because evidence failed. It lost because scientists stayed cautious while nonsense went viral. Fear became entertainment. Confidence replaced expertise. Scientists write careful paragraphs, influencers get millions of views with one emotional lie.
75
149
7
642
10,110
Peter Adams retweeted
Have a look at the first row of top videos that come up on TikTok when I simply search for "hantavirus" on the platform.
9
24
3
99
15,447
The collapse of trust in science is going to go down in history as one of the most sad, bizarre, and destructive social contagions of modern times. We fed billions, cured diseases and powered nations - yet people ran toward conspiracies instead.
5,927
5,307
748
24,658
1,972,413
New: McClatchy journalists at newspapers in 4 states are now withholding their bylines from A.I.-generated content as tensions grow over a "content scaling agent" tool that the company rolled out nytimes.com/2026/05/01/busin…
8
489
50
944
409,187
Peter Adams retweeted
Please read our incredible media writer @dbauder's final piece in his illustrious career at the @AP: A photographer has captured some 50 newsrooms to document places and lives endangered by the industry’s collapse apnews.com/article/newspaper…
9
423
21
794
129,350
Peter Adams retweeted
I am among the McClatchy journalists withholding my byline from AI-generated content. I cover sensitive stories that should never be repackaged or altered in any way by artificial intelligence. Whether the story is about sexual abuse or immigrants caged at Alligator Alcatraz, our journalists take great care to ensure that their material is accurate, fair and conveys invaluable aspects of humanity that AI simply can't do.
65
1,508
26
5,821
253,366
The web is disappearing 🕳️ According to a Pew Research Center report, 26% of pages from 2013-2023 are no longer accessible. But that’s not the whole story. In a new study published in Internet Archive's book, VANISHING CULTURE, data scientists working with the Wayback Machine have found: 16% have been restored through the Wayback Machine. 56% are preserved before they disappear. Preservation is the remedy for cultural loss. 📚 Read VANISHING CULTURE free from the Internet Archive 📖 Download & read: archive.org/details/vanishin… 🛒 Purchase in print: betterworldbooks.com/product… #VanishingCulture #DigitalMemory #InternetArchive #BookTwitter
209
4,314
252
12,389
571,080
Peter Adams retweeted
“In the longer term, the greatest danger posed by AI isn’t fake images themselves, it’s the collapse of trust in real evidence,” writes Poynter contributor Adam Rose. “Seeing is no longer believing — and our institutions are unprepared for it.” Here’s why journalism needs to prove which images are authentic — not just label deepfakes: poynter.org/commentary/2026/…
6
5
952
Peter Adams retweeted
New data suggests small publishers (10k daily pageviews or less) are hit hardest by search traffic declines —Chatbots only drive 1% of all traffic referrals, so they can't make up the gap —News sites get most page views from AI platforms, but the lowest engagement @axios axios.com/2026/03/17/chartbe…
1
50
4
90
31,322
This is the work that can’t be replicated by AI or pursued by a reporter with a Substack. It requires time - during which a reporter isn’t producing a lot of copy. Building trust with sources so you can tell sensitive stories, that’s a process that doesn’t scale.
"Nearly five years." Investigative journalism takes years and years. The investment is immense, and always worth it, and, in the era of "hot takes" and ephemeral "content," rigorously reported journalistic work is needed more than ever. We must keep supporting investigative journalism. From @mannyNYT, lead reporter of the blockbuster piece on Cesar Chavez ⤵️
4
312
5
1,705
75,100
Peter Adams retweeted
This is totally insane. A war correspondent just received death threats from online gamblers who wanted him to change his reporting on an Iranian missile strike so they could collect a payout. One bettor had $900,000 riding on the outcome. He told the journalist he knew where he lived and who his family members were.  This is what prediction markets on life and death actually look like in practice. This is exactly why I introduced the DEATH BETS Act with Senator @AdamSchiff. The DEATH BETS Act would ban contracts on assassinations, deaths of world leaders, and acts of war on platforms like Polymarket. This story shows exactly why that matters.  When you let people place million-dollar bets on whether a missile kills someone, you create a financial incentive to threaten journalists, manipulate information, and profit from human suffering. washingtonpost.com/technolog…
345
4,960
362
13,674
923,678
This video posted by the official Persian account of the Israeli government claims to show a member of the Basij paramilitary force on the streets of Tehran being targeted by a strike. But it's AI-generated. Follow this thread to find out how we can verify that it's AI.
ما شرایط را برای شما فراهم می‌کنیم. قدم به قدم.
12
161
18
530
99,695