Erik Sofge (@esofge) header
Erik Sofge (@esofge) profile picture

Senior Editor at MIT Horizon. Used to write for Popular Science, Men’s Journal, Slate, Facebook, etc.

Massachusetts
Joined October 2010
Photo by Erik Sofge (@esofge)Photo by Erik Sofge (@esofge)Photo by Erik Sofge (@esofge)Photo by Erik Sofge (@esofge)Photo by Erik Sofge (@esofge)Photo by Erik Sofge (@esofge)Photo by Erik Sofge (@esofge)Photo by Erik Sofge (@esofge)Photo by Erik Sofge (@esofge)Photo by Erik Sofge (@esofge)
Erik Sofge retweeted
People want to fight fire with fire… but this isn't the answer. The #hiring process is already overwhelmed with #AI's presence as it tries to adapt – candidates need a system that improves their chances instead of ruining them. @JPROCO inc.com/joe-procopio/job-see…
1
3
193
Erik Sofge retweeted
Google está muerto.
Boomers told us not to trust wikipedia only to fall for this
133
4,952
1,364
39,401
14,229,374
“I have invented the volleyball that Tom Hanks talks to when he’s marooned for 8 years.”
This quoted post is unavailable.
16
67
1
503
19,375
o1 prompting is alien to me. Its thinking, gloriously effective at times, is also dreamlike and unamenable to advice. Just say what you want and pray. Any notes on “how” will be followed with the diligence of a brilliant intern on ketamine.
I ask o1-mini to write a poem about strawberries with 13 R’s, and use five separate tests to verify its answer. It returns a poem with 21 R’s, and details passing each of five hallucinated checks like it’s a human at a computer: “I applied a regex,” “I read the poem aloud.”
17
22
10
383
54,299
I am begging you to listen to the cyberpunk backing music in the home office's announcement they're banning katanas
🚨 Ninja swords are used by criminals to rob victims of their lives. There is only one course of action: ✅ Remove knives from the streets ✅ Tougher consequences for criminals ✅ Save lives
208
1,923
31
17,498
775,582
The AI hype cycle is damaging real innovation across tech. It is crazy FOMO having negative impact on the world as we move forward. "AI is artificially inflating valuation numbers across venture: The segment captured nearly 50% of all deal value in Q2. But flat and down rounds are at a decade high and valuation step-ups between rounds have softened, according to PitchBook’s Q2 2024 US VC Valuations Report, sponsored by Morgan Stanley at Work and Forge Global." pitchbook.com/news/reports/q…
10
35
5
159
52,544
the teachers use ChatGPT to create the assignments the students use ChatGPT to complete the assignments the teachers use ChatGPT to grade the assignments why is anybody even showing up
A great piece which correctly points out that the AI cheating crisis should be understood in the context of the broader academic labor crisis, and that it’s not just students, but also overworked faculty, who are resorting to chatbots to cut down on their work. Dismal stuff.
70
2,149
112
10,637
447,790
Attempting to make a home loan payment on X: The Everything App and accidentally sending the word "peñus" to the official account of the Milwaukee Brewers.
26
626
20
5,461
394,256
After months of digging and reporting, I have learned where Facebook's bizarre AI spam (like "Shrimp Jesus") comes from, who is making it, how it works, and how it is monetized. Turns out Meta is directly paying people to spam FB with this stuff 404media.co/where-facebooks-…
59
2,005
250
5,678
968,754
Because the government came & took all my computers away, we had to improvise. Here's the video I promised about how I used a 2003 eMac I found on the side of the road to capture and edit HD broadcast video, in 2023:
12
46
5
431
302,077
"Whether a person or a program wrote the posts does not matter, really; the quality of the product is the same, and perfectly reflects the disregard that The Arena Group has for both its readers and the people at Sports Illustrated." defector.com/scorn-illustrat…
2
23
5
86
45,108
It’s true! The people who seem most excited about AI are the people who care the least about the quality of the work! Feels like that’s not a coincidence!
This Post is from an account that no longer exists. Learn more
4
47
2
387
33,785
Not sure if there's a person on Earth who's better at their job than Matt Levine. Here, he sums up the OpenAI fiasco in two charts (i.e. the board nominally controls the company, but can't actually run the thing without Microsoft's cash) bloomberg.com/opinion/articl…
4
51
5
281
73,851
Recent clip of @saranormous chatting with @ilyasut has some interesting stuff on his thoughts on alignment. It hardly sounds like he's seems some shocking new capabilities. More likely he sees little progress in "superalignment" so far youtu.be/Ft0gTO2K85A?si=DQBg…
2
1
8
2,077
This is the best example I've found yet of how derivative AI 'art' is. The person who generated the image on the left asked Midjourney to generate 'an average woman of Afghanistan'. It produced an almost carbon copy of the 1984 photo of Sharbat Gula, taken by Steve McCurry. 1/
184
7,320
396
68,666
5,002,299
Here's how generative AI will take over journalism: publishers will use it while claiming that the material "was created by third-party freelancers hired by a marketing agency partner, not AI.” nypost.com/2023/10/27/media/…
1
1
2
179
Absolutely wild that a major newspaper appears to be using AI to create articles and publishing them to their website under bylines of writers that do not appear to exist. What did mgmt think was going to happen when these articles just started appearing? washingtonpost.com/style/med…
22
232
41
693
117,528
When people in 2033 look back at 2023 AI the way we look back on early cell phones
21
13
6
68
10,127
People arguing for AI rights based on complex text processing algorithms need to ask whether they would assign the same rights to calculators, smart watches, and the internet. “I don’t quite get it how works” + “it surprises me” ≠ it could maybe be sentient if I squint.
My new essay @TheHill argues we need an AI rights movement. AIs are no longer just tools. They are quickly becoming digital minds integrated in society as friends and coworkers. The future turns on whether and how we include them in the moral circle. 🧵thehill.com/opinion/cybersec…
65
83
13
473
98,243