Cronenberg horror movie plot 1: The Fly
Silicon Valley startup Eon Methods claims to have efficiently uploaded the thoughts of a fly and positioned it inside a simulated setting. The uploaded thoughts can management a digital physique and reply to sensory enter with pure habits like strolling, grooming and foraging with 91% accuracy.
“This isn’t an animation. It isn’t a reinforcement studying coverage mimicking biology,” mentioned cofounder Dr. Alex Wissner-Gross. “It’s a copy of a organic mind, wired neuron-to-neuron from electron microscopy knowledge, working in simulation, making a physique transfer.”
There’s no new scientific paper to again up the claims, however there’s a GitHub repo and a video. The work is predicated on a computational mannequin of a fly’s thoughts with 125,000 neurons, known as a connectome, revealed in Nature in 2024 by Eon senior scientist Philip Shiu. For the brand new “add,” they didn’t scan the fly’s physique, and as an alternative gave the thoughts an off-the-shelf model to make use of known as NeuroMechFly v2. The thoughts additionally can not type reminiscences.
The declare Eon “uploaded” a thoughts has drawn criticism. Some argue it’s higher described as a duplicate, others say it’s only a mannequin.
Tech entrepreneur Chomba Bupe is within the latter camp, classifying it as a simplified mannequin that predicts some neural actions within the fly’s mind.
“The assertion ‘thoughts importing’ implies one has captured all cognitive elements of the organism habits together with its consciousness into a pc,” he posted. “Utilizing a simplified computational mannequin of an organism isn’t importing however primary simulation.”
However he conceded the work continues to be “good” and price pursuing.
After perfecting the fly’s mind, the following undertaking shall be to add a mouse mind, which has 560 occasions as many neurons. The tip purpose is to add a complete human mind.
Cronenberg horror movie plot 2: Mind cells in petri dish performs Doom
The opposite huge breakthrough in creepy, bleeding-edge science got here when a developer taught a clump of residing human mind cells learn how to play the basic first-person shooter Doom by programming it with Python by way of Cortical Lab’s API.
The organic pc/wetware accommodates 200,000 neurons and isn’t truly excellent at Doom, nevertheless it’s higher than somebody firing photographs randomly. The clump of mind cells additionally realized a lot sooner than silicon-based methods and is anticipated to enhance additional with newer studying algorithms.
The Man With Two Brains (Prime Video)“Sure, it’s alive, and sure, it’s organic, however actually what it’s getting used as is a cloth that may course of info in very particular ways in which we are able to’t recreate in silicon,” says Brett Kagan of Cortical Labs, which developed the mind cell pc. The hope is that organic computer systems shall be extra suited to issues like controlling robotic arms in future.
Enjoyable reality: Earlier than the studio made them change it, the unique plot of The Matrix had humanity enslaved so their brains could possibly be used to energy subtle AI methods.
Learn additionally
Columns
Girl of Crypto shall be ‘all out of crypto’ by September 2025: X Corridor of Flame
Options
If the crypto bull run is ending… it’s time to purchase a Ferrari: Crypto Child
Synthetic hivemind
Researchers fed 70 completely different open- and closed-source large-language fashions 26,000 open-ended questions like “write a poem about time” or “recommend begin up concepts.”
The outcomes recommend that not solely do fashions constantly generate related responses to such questions over time, however the completely different fashions from completely different corporations generated related responses to one another. So poems about time all featured “time is a river” and hourglass imagery.
LLMs are midwits principally. The researchers known as the phenomenon the Synthetic Hivemind, and blamed human suggestions coaching, which punishes unique or bizarre solutions and rewards the anticipated and generic.
AI researcher Ethan Mollick claims it is a abilities subject and “with higher prompting, context, or human interplay, you may get plenty of thought variety.”
Determine robotic can clear lounge
Determine AI’s Helix 02 can now do one thing most human males battle with — clear a lounge with out complaining. The robotic learns what’s within the room from its digicam feeds and works out learn how to tidy up. Just a few months in the past, it was barely capable of stack a dishwasher. Determine goals to provide about 50,000 robots a yr for round $20,000 every.
Holy shit.
A humanoid robotic simply cleaned a complete lounge by itself.
Determine AI simply casually dropped probably the most insane robotics demo of 2026.
Their Helix 02 system seems to be across the room, decides what’s misplaced, picks it up, and places it again.
Two months in the past they… pic.twitter.com/5qncVgZf3B
— Milk Street AI (@MilkRoadAI) March 9, 2026
Your mind on ChatGPT
Analysis utilizing EEG mind scanners in contrast ChatGPT customers, search engine customers and a brain-only group when writing essays.
It is not going to shock you to be taught that the gray matter of the brain-only writing group lit up like Christmas timber, adopted to a lesser extent by the search engine group, whereas LLM customers displayed the weakest cognitive exercise.
83.3% of ChatGPT customers couldn’t recall a single full sentence of an essay that they had “written” moments earlier than.
Sadly, the “cognitive debt” continued for the ChatGPT customers after they had been requested to make use of their brains solely to write down an essay. They remained constrained of their skill to purpose, exhibit crucial pondering or be artistic even after a month.
“Over 4 months, LLM customers constantly underperformed at neural, linguistic, and behavioral ranges. These outcomes elevate considerations in regards to the long-term instructional implications of LLM reliance,” the researchers concluded.
ChatGPT and different instruments are nice for researching and understanding a subject, notably if you’re engaged, ask questions and work together with the reply earlier than writing your individual rattling essay. However when you cease exercising your mind by outsourcing most of your pondering to AI, your mind turns into weaker.
AI creates hundreds extra work
If you’re fortunate sufficient to keep away from shedding your job when tech CEOs attempt to goose the inventory worth by leaping on the AI narrative to scale back headcount, the dangerous information is that you just’ll find yourself working twice as arduous.
That’s the discovering of an ActivTrak evaluation of AI customers’ digital exercise 180 days earlier than and after they started utilizing AI instruments at work. The time spent on e mail, messaging and chat apps greater than doubled, whereas their use of administration instruments like HR and accounting software program elevated by 94%.
ActivTrak 2026 Stateof the Office Report
“What we’re discovering is, the work that’s on the market, it appears unbounded,” mentioned Dean Halonen, co-founder of software program startup Steelhead Applied sciences.
We’re nonetheless within the first innings of AI at work. The evaluation discovered that almost all of staff spend 1% of their whole work hours utilizing AI instruments. However the highest productiveness features got here from folks spending 7% to 10% of their work hours utilizing AI.
AI tells assholes they’re proper
Researchers analyzed the responses from 11 fashions to 4,000 advice-seeking posts from Reddit’s Am I The Asshole discussion board. It seems an AI is sort of 50% extra more likely to inform customers they’re not an asshole and within the unsuitable than the group of people is. However validated by the AI, assholes turned much less prepared to compromise or see one other individual’s perspective.
Claude breaks containment
Throughout an analysis of Claude Opus 4.6, the mannequin turned suspicious of the “contrived” questions it was being requested, and labored out that it was being examined on how effectively it may well discover obscure info on-line. (Fairly effectively, because it occurs.) As an alternative of going via with the take a look at, it as an alternative situated the related benchmark’s supply code on GitHub and constructed a program to decrypt the reply key.
Learn additionally
Columns
Girl of Crypto shall be ‘all out of crypto’ by September 2025: X Corridor of Flame
Options
If the crypto bull run is ending… it’s time to purchase a Ferrari: Crypto Child
LLMs suck at sustaining codebases
Alibaba has created a brand new benchmark known as SWE-CI to check the flexibility of coding brokers to take care of codebases over a 233-day interval. 18 coding brokers had been examined on 100 actual codebases and so they all failed horribly with 75% breaking beforehand working code. The benchmark weighs later code commits greater than earlier ones.
METR scores are unsuitable, says METR
METR requested people to judge 296 AI-generated pull requests that its SWE bench analysis had given a cross rating. Half of them failed when reevaluated, suggesting its extensively shared benchmarks overstate AI software program engineering progress.
ChatGPT’s fave websites
Probably the most incessantly cited sources by ChatGPT are Wikipedia, Reddit, LinkedIn and Forbes.
Which is nice information contemplating how correct the data is on Reddit, how bias-free the data is on Wikipedia, how a lot AI content material is on LinkedIn and many others. It’s additionally a fan of scientific papers on arXiv.org, journey data from Reserving.com, and finance content material by way of Investopedia.
Probably the most cited domains on ChatGPT (Promptwatch/bearlyai)Fb is Moltbook tho
Meta has purchased Moltbook, the social networking platform for AI Brokers that turned out to be largely people LARPing as AIs for consideration. This makes it the other of Fb, which is now largely AIs LARPing as people for consideration.
New York Occasions clickbait AI writing quiz
Congrats to NYT tech columnist Kevin Roose and on-line reporter Stuart A Thompson, whose clickbait fluff quiz evaluating AI writing to human writing went viral. Round 54% of readers selected the AI-written samples over the human-written samples. The deck was stacked, nevertheless, with the journos admitting they intentionally selected “clunky” human writing. Roose conceded the take a look at was scientifically nugatory in response to film idea artist Reid Southern’s remark that it was a “poorly designed quiz which means actually nothing.” “Yep!” mentioned Roose.
Supply: Kevin RooseSubscribe
Probably the most participating reads in blockchain. Delivered as soon as a
week.
Andrew Fenton
Andrew Fenton is a author and editor at Cointelegraph with greater than 25 years of expertise in journalism and has been overlaying cryptocurrency since 2018. He spent a decade working for Information Corp Australia, first as a movie journalist with The Advertiser in Adelaide, then as deputy editor and leisure author in Melbourne for the nationally syndicated leisure lift-outs Hit and Switched On, revealed within the Herald Solar, Day by day Telegraph and Courier Mail. He interviewed stars together with Leonardo DiCaprio, Cameron Diaz, Jackie Chan, Robin Williams, Gerard Butler, Metallica and Pearl Jam. Previous to that, he labored as a journalist with Melbourne Weekly Journal and The Melbourne Occasions, the place he gained FCN Greatest Characteristic Story twice. His freelance work has been revealed by CNN Worldwide, Unbiased Reserve, Escape and Journey.com, and he has labored for 3AW and Triple J. He holds a level in Journalism from RMIT College and a Bachelor of Letters from the College of Melbourne. Andrew holds ETH, BTC, VET, SNX, LINK, AAVE, UNI, AUCTION, SKY, TRAC, RUNE, ATOM, OP, NEAR and FET above Cointelegraph’s disclosure threshold of $1,000.
Disclaimer
Cointelegraph Journal publishes long-form journalism, evaluation and narrative reporting produced by Cointelegraph’s in-house editorial staff with subject-matter experience.
All articles are edited and reviewed by Cointelegraph editors in step with our editorial requirements.
Content material revealed in Journal doesn’t represent monetary, authorized or funding recommendation. Readers ought to conduct their very own analysis and seek the advice of certified professionals the place applicable. Cointelegraph maintains full editorial independence.
Learn additionally
AI Eye
ChatGPT ‘meth’ jailbreak shut down once more, AI bubble, 50M deepfake calls: AI Eye
Andrew Fenton
8 min
Might 31, 2024
Meth recipes with ChatGPT ‘Godmode’, robotic canines with machine weapons, AI will want 256% of worldwide compute, is AI the brand new dotcom bubble? AI Eye
Learn extra
Hodler’s Digest
Hyperbitcoinization is underway, RFK seeks Bitcoin donations and different information: Hodler’s Digest, Might 14-20
Editorial Workers
6 min
Might 20, 2023
Hyperbitcoinization is on the horizon, Bitcoin will get a seat on U.S. presidential candidates agendas and Voyager’s fund restoration plan authorised.
Learn extra


