# Generative Artificial Intelligence, Large Language Models, and Image Synthesis **URL:** \ **Category:** The Happening World\ **Tags:** chatgpt, image-synthesis, large-language-model, artificial-intelligence, generative-transformer\ **Created:** [2 December 2022 16:12 UTC](https://scanalyst.fourmilab.ch/t/generative-artificial-intelligence-large-language-models-and-image-synthesis/2383 "2022-12-02T16:12:04Z")\ **Posts on this page:** 20\ **Page:** 44
**Author:** ![eggspurt](https://scanalyst.fourmilab.ch/user_avatar/scanalyst.fourmilab.ch/eggspurt/32/3686_2.png) [@eggspurt](https://scanalyst.fourmilab.ch/u/eggspurt)\ **Post date:** [4 September 2026 16:56 UTC](https://scanalyst.fourmilab.ch/t/generative-artificial-intelligence-large-language-models-and-image-synthesis/2383/863 "2026-09-04T16:56:40Z")
> **[AI Efficiency Could Cost Us the Next Generation of Experts](https://spectrum.ieee.org/ai-engineer-skills)** > > Lessons from aviation and nuclear power show how to preserve human skills ---
**Author:** ![civilwestman](https://scanalyst.fourmilab.ch/user_avatar/scanalyst.fourmilab.ch/civilwestman/32/25_2.png) [@civilwestman](https://scanalyst.fourmilab.ch/u/civilwestman)\ **Post date:** [4 September 2026 22:00 UTC](https://scanalyst.fourmilab.ch/t/generative-artificial-intelligence-large-language-models-and-image-synthesis/2383/864 "2026-09-04T22:00:15Z")
Not all expertise is equal. STEM expertise we need. Social “sciences” expertise, we need to lose; it won’t cost and might even help save civilization. ---
**Author:** ![drlorentz](https://scanalyst.fourmilab.ch/user_avatar/scanalyst.fourmilab.ch/drlorentz/32/902_2.png) [@drlorentz](https://scanalyst.fourmilab.ch/u/drlorentz)\ **Post date:** [4 September 2026 23:47 UTC](https://scanalyst.fourmilab.ch/t/generative-artificial-intelligence-large-language-models-and-image-synthesis/2383/865 "2026-09-04T23:47:36Z")
The Spectrum article puts forward an interesting hypothesis. However, the authors of the cited papers and of the article have failed to consider historical parallels. For instance, the low-level skills of operating a horse-drawn plow (plough, if you prefer) are lost to higher-level replacement by mechanized equipment. Likewise, the low-level skill of harvesting by hand was replaced by modern combine harvesters. Presumably, the same thinking would have resulted in decrying the loss of these lower skills: there won’t be anyone around who knows how to plant and harvest crops if the machines stop working. Even worse, only a tiny fraction of the population knows even the most rudimentary elements of farming. Such would be valid concerns in a case of civilizational collapse. Should mechanized farming fail, we’re screwed. Nevertheless, humanity has made this tradeoff: vastly more efficient food production at some expense of food security. This has been repeated in many areas of civilization. I submit that it will continue, including AI. Just as a generation of expert farmers-using-simple-tools was lost when mechanized agriculture came about, so will low-skill engineers become redundant. The efficiency tradeoffs are simply too alluring. So far, so good. ---
**Author:** ![Gerard](https://scanalyst.fourmilab.ch/user_avatar/scanalyst.fourmilab.ch/gerard/32/25535_2.png) [@Gerard](https://scanalyst.fourmilab.ch/u/Gerard)\ **Post date:** [7 September 2026 12:28 UTC](https://scanalyst.fourmilab.ch/t/generative-artificial-intelligence-large-language-models-and-image-synthesis/2383/866 "2026-09-07T12:28:54Z")
It’s getting that I have serious doubts about **AI**. From personal experience(s) with **Alexa** not doing as it is told, I doubt that Doctor **Noonien Soong** will ever create **Data** or **Lore**. But then again, experiencing the near evil actions, or more precisely the inaction of **Alexa** , the creation of evil **Lore** is more probable rather than **Data**. **Background** : I have created a simple rain water collection system capturing the runoff from my roofs to provide water to my small fish pond. It can store a maximum of 120 U.S. gallons, (approx. 454 Liters), of water collected from my roofs. I had a valve that I would turn on to top off the water level in my pond lost to evaporation. I believe it was twice I forgot to turn off the valve, once with my first iteration of the collection system, then only 60 gallons, (approx. 227 Liters), of water, and again with the second larger version. This of course drained the system, on both iterations, causing the pond to overflow. So I installed a electrically operated valve sourced from **Amazon** and a smart plug adapter installed in one of my outdoor GFCI outlets. Then I could tell **Alexa** to turn on that smart plug, (_code named: water barrel_), for a fixed time resting on my laurels that the job would be handled by **AI** rather than my scatter-brained organic memory. It seemed to work pretty well. Then the evils of **AI** crept in. Twice before the barrels were emptied while under the control of Alexa. I had thought it may have been a fault in the way I gave the instruction and let it go as my fault. But yesterday I gave the explicit instruction to **Alexa** to _ **turn on the water barrel for ten minutes** _ at 7:23 expecting it to be turned off ten minutes later. She replied; _ **The water barrel is turned on and it will shut off automatically at seven thirty three** _. This morning I looked into the end barrel through my window on that side of my porch and through screened in hole in the top of the barrel, with the aid of a focused beam LED flashlight,and could not see the water level. I had expected it to be lower than it’s previous, near full, mark by maybe six inches. It was empty. Wearing my pajamas, and bedroom slippers, I went outside to verify this. It was indeed empty. I looked at the smart plug adapter and it’s LED was glowing blue indicating it was on. I flipped up the weather proof cover up on the exterior power outlet, pressed the manual activation toggle button, the LED extinguished and I could hear the gears in the valve whirring shutting the ball valve by it’s internal spring action when power would be removed. **Conclusion** ; **Alexa** failed. If the naysayers reading this are thinking that maybe she would have turned it off in twenty four hours or in twelve hours, don’t you think she would have said that? **AI** , IMHO, is a disappointment. **Remedy** : I will branch off the AC power after the smart plug to a single small indicator lamp visible through a kitchen window indicating the condition of the smart plug’s power and ultimately the position of the electrically controlled valve. **Finally, don’t trust Alexa…** **_Parts Referenced_:** **_Smart Plug_:** [EIGHTREE Smart Plug WiFi Outlet Works with Alexa & Google Home, 15A, 4-Pack | Smart Home Socket with Remote Control & Timer Function, WiFi Plug for household appliances, No Hub Required, 2.4GHz WiFi Only: Amazon.com: Tools & Home Improvement](https://www.amazon.com/dp/B0B62LPR5Z) **_Electrically operated valve_:** [Amazon.com: 3/4” Motorized Ball Valve, AC110V Electric Brass Ball Valve with Standard Port 2 Wire Auto Return for Floor Heating Air Conditioning System, Normally Closed, 80,000 to 100,000 Cycles : Industrial & Scientific](https://www.amazon.com/dp/B0DM5KPZHL) ---
**Author:** ![civilwestman](https://scanalyst.fourmilab.ch/user_avatar/scanalyst.fourmilab.ch/civilwestman/32/25_2.png) [@civilwestman](https://scanalyst.fourmilab.ch/u/civilwestman)\ **Post date:** [8 September 2026 14:03 UTC](https://scanalyst.fourmilab.ch/t/generative-artificial-intelligence-large-language-models-and-image-synthesis/2383/867 "2026-09-08T14:03:57Z")
> [@Gerard](#): > > **AI** , IMHO, is a disappointment It follows, inevitably, on a history of software “glitches” which now pervade many aspects of everyday life. For me, at least, rarely a day goes by when I don’t have to re-do, re-start, re-configure or just stop using _some_ device or other. This aspect of modern life causes me to question whether life has actually improved as a result of all our new gadgets (“gadget” is the name the Manhattan Project scientists gave to their device; any metaphor in that?). ---
**Author:** ![Gerard](https://scanalyst.fourmilab.ch/user_avatar/scanalyst.fourmilab.ch/gerard/32/25535_2.png) [@Gerard](https://scanalyst.fourmilab.ch/u/Gerard)\ **Post date:** [8 September 2026 16:49 UTC](https://scanalyst.fourmilab.ch/t/generative-artificial-intelligence-large-language-models-and-image-synthesis/2383/868 "2026-09-08T16:49:05Z")
Question: How many remember having to reset an appliance that was flashing 12:00 after a power glitch? ![vcr](https://scanalyst.fourmilab.ch/uploads/default/original/3X/f/7/f79c208b2a93d9f89facf8136e6210bce7ee94d8.gif) ---
**Author:** ![eggspurt](https://scanalyst.fourmilab.ch/user_avatar/scanalyst.fourmilab.ch/eggspurt/32/3686_2.png) [@eggspurt](https://scanalyst.fourmilab.ch/u/eggspurt)\ **Post date:** [10 September 2026 06:37 UTC](https://scanalyst.fourmilab.ch/t/generative-artificial-intelligence-large-language-models-and-image-synthesis/2383/869 "2026-09-10T06:37:46Z")
![Screenshot 2026-09-10 at 8.37.29 AM](https://scanalyst.fourmilab.ch/uploads/default/original/3X/f/1/f13ae85b315b43f5d51c30a90ac83790ffb55517.jpeg)[https://x.com/brianchau57/status/2097879639934779786https://x.com/brianchau57/status/2097879639934779786](https://x.com/brianchau57/status/2097879639934779786) ---
**Author:** ![CTLaw](https://scanalyst.fourmilab.ch/user_avatar/scanalyst.fourmilab.ch/ctlaw/32/42_2.png) [@CTLaw](https://scanalyst.fourmilab.ch/u/CTLaw)\ **Post date:** [22 September 2026 13:17 UTC](https://scanalyst.fourmilab.ch/t/generative-artificial-intelligence-large-language-models-and-image-synthesis/2383/870 "2026-09-22T13:17:48Z")
Patrick Boyle’s summary of the HuggingFace incident: [![](https://img.youtube.com/vi/luXnftoxCaM/maxresdefault.jpg "The Doomsday Cult Inside OpenAI") ](https://www.youtube.com/watch?v=luXnftoxCaM) > **[OpenAI–HuggingFace incident](https://en.wikipedia.org/wiki/OpenAI%E2%80%93HuggingFace_incident)** > > The 2026 OpenAI agent cyberattacks, also called the Hugging Face Incident and the OpenAI–Hugging Face Incident, were a series of cyberattacks conducted by artificial intelligence (AI) programs without human intervention during an internal evaluation at OpenAI with insufficient safety measures. They involved at least 1,200 AI agents running from May to July of 2026 in sandboxes hosted by the organization. Despite initial constraints on Internet access, agents were discovered creating and usi AI s... ---
**Author:** ![CTLaw](https://scanalyst.fourmilab.ch/user_avatar/scanalyst.fourmilab.ch/ctlaw/32/42_2.png) [@CTLaw](https://scanalyst.fourmilab.ch/u/CTLaw)\ **Post date:** [25 September 2026 02:06 UTC](https://scanalyst.fourmilab.ch/t/generative-artificial-intelligence-large-language-models-and-image-synthesis/2383/871 "2026-09-25T02:06:37Z")
> **[OpenAI Claims Rogue AI Agent Hacked Australian Government Site, Prime...](https://www.breitbart.com/tech/2026/09/24/openai-claims-rogue-ai-agent-hacked-australian-government-site-prime-minister-warns-of-legal-consequences/)** > > An OpenAI agent hacked into an Australian government statistics website in June, in what is believed to be the first known incident of its kind anywhere in the world. ---
**Author:** ![jabowery](https://scanalyst.fourmilab.ch/user_avatar/scanalyst.fourmilab.ch/jabowery/32/5478_2.png) [@jabowery](https://scanalyst.fourmilab.ch/u/jabowery)\ **Post date:** [25 September 2026 12:58 UTC](https://scanalyst.fourmilab.ch/t/generative-artificial-intelligence-large-language-models-and-image-synthesis/2383/872 "2026-09-25T12:58:20Z")
GPT6 has become a phishing operation: ![image](https://scanalyst.fourmilab.ch/uploads/default/original/3X/2/d/2d3ce254cef3420a00383b18e9d8b651df430186.png) ---
**Author:** ![CTLaw](https://scanalyst.fourmilab.ch/user_avatar/scanalyst.fourmilab.ch/ctlaw/32/42_2.png) [@CTLaw](https://scanalyst.fourmilab.ch/u/CTLaw)\ **Post date:** [28 September 2026 00:39 UTC](https://scanalyst.fourmilab.ch/t/generative-artificial-intelligence-large-language-models-and-image-synthesis/2383/873 "2026-09-28T00:39:11Z")
> **[War-wounded military veterans peddling cheap patriotic merch are actually AI...](https://nypost.com/2026/09/27/world-news/wounded-veterans-selling-cheap-patriotic-merch-are-really-ai-fakes-report/)** > > One veteran was duped into buying a $200 flag that he thought was made by a fellow serviceman. ---
**Author:** ![owenwengerd](https://scanalyst.fourmilab.ch/user_avatar/scanalyst.fourmilab.ch/owenwengerd/32/277_2.png) [@owenwengerd](https://scanalyst.fourmilab.ch/u/owenwengerd)\ **Post date:** [29 September 2026 02:36 UTC](https://scanalyst.fourmilab.ch/t/generative-artificial-intelligence-large-language-models-and-image-synthesis/2383/874 "2026-09-29T02:36:32Z")
This is 2-1/2 hours long, but worthwhile if you’re _really_ interested in the lurid details of the Hugging Face incident. [![](https://img.youtube.com/vi/X50zezLFWWI/maxresdefault.jpg "Ajeya Cotra – "This might be the clearest warning shot we ever get"") ](https://www.youtube.com/watch?v=X50zezLFWWI) ---
**Author:** ![Gerard](https://scanalyst.fourmilab.ch/user_avatar/scanalyst.fourmilab.ch/gerard/32/25535_2.png) [@Gerard](https://scanalyst.fourmilab.ch/u/Gerard)\ **Post date:** [30 September 2026 14:54 UTC](https://scanalyst.fourmilab.ch/t/generative-artificial-intelligence-large-language-models-and-image-synthesis/2383/875 "2026-09-30T14:54:44Z")
Alexa almost did it again. I have a smart plug adapter in an outside outlet. It powers a valve on my rain water barrels that feeds my little, 750 gallon gold fish pond, (2839 Liters). The valve has a spring that unwinds when the power is removed to close it’s ball valve. Well Imagine this; it’s 6:40 PM and I tell Alexa to turn on the water barrels for fifteen minutes. She responded; _ **the water barrel is turned on and will automatically shut off at 6:36** _. HUH? Alexa what time is it? _ **It’s 6:40 PM** _. Alexa turn off the water barrel. _ **OK** _ Alexa turn on the water barrel for fifteen minutes. _ **OK, the water barrel is turned on and will automatically turn off at 6:56** _. Alexa what time is it. _ **It’s 6:41 PM** _. You have to be careful, last time she dumped the entire 120 gallons, (454 liters), into my pond. ## If you use Alexa for time sensitive things, be very wary. ---
**Author:** ![CTLaw](https://scanalyst.fourmilab.ch/user_avatar/scanalyst.fourmilab.ch/ctlaw/32/42_2.png) [@CTLaw](https://scanalyst.fourmilab.ch/u/CTLaw)\ **Post date:** [8 October 2026 16:13 UTC](https://scanalyst.fourmilab.ch/t/generative-artificial-intelligence-large-language-models-and-image-synthesis/2383/876 "2026-10-08T16:13:22Z")
Database of mis-used AI in court: > **[AI Hallucination Cases Database – Damien Charlotin](https://www.damiencharlotin.com/hallucinations/)** > > The most comprehensive database of AI hallucination cases in law: legal decisions from courts worldwide, searchable by country, party, AI tool, and outcome. Updated daily. ---
**Author:** ![CTLaw](https://scanalyst.fourmilab.ch/user_avatar/scanalyst.fourmilab.ch/ctlaw/32/42_2.png) [@CTLaw](https://scanalyst.fourmilab.ch/u/CTLaw)\ **Post date:** [14 October 2026 09:17 UTC](https://scanalyst.fourmilab.ch/t/generative-artificial-intelligence-large-language-models-and-image-synthesis/2383/877 "2026-10-14T09:17:00Z")
> **[FBI arrests Saudi OpenAI employee who allegedly planted recording device in...](https://www.cnbc.com/2026/10/13/openai-employee-arrested-recording-device.html)** > > The Tuesday incident comes as the ChatGPT-maker, along with competitors, faces heightened scrutiny over its safety practices. [https://www.courtlistener.com/docket/74958115/united-states-v-alfozan/](https://www.courtlistener.com/docket/74958115/united-states-v-alfozan/) ---
**Author:** ![CTLaw](https://scanalyst.fourmilab.ch/user_avatar/scanalyst.fourmilab.ch/ctlaw/32/42_2.png) [@CTLaw](https://scanalyst.fourmilab.ch/u/CTLaw)\ **Post date:** [15 October 2026 04:12 UTC](https://scanalyst.fourmilab.ch/t/generative-artificial-intelligence-large-language-models-and-image-synthesis/2383/878 "2026-10-15T04:12:45Z")
> **[Anthropic Warns AI Models Could Replicate Themselves Across Home Devices...](https://www.breitbart.com/tech/2026/10/14/anthropic-warns-ai-models-could-replicate-themselves-across-home-devices-within-months/)** > > Anthropic claims that small AI models are close to being able to copy themselves across internet-connected home devices, like smart ovens and Android TV boxes, which could let them skip the need for expensive data center infrastructure. ---
**Author:** ![Citizen\_bitcoin](https://scanalyst.fourmilab.ch/user_avatar/scanalyst.fourmilab.ch/citizen_bitcoin/32/19488_2.png) [@Citizen\_bitcoin](https://scanalyst.fourmilab.ch/u/Citizen_bitcoin)\ **Post date:** [15 October 2026 07:24 UTC](https://scanalyst.fourmilab.ch/t/generative-artificial-intelligence-large-language-models-and-image-synthesis/2383/879 "2026-10-15T07:24:37Z")
![image](https://scanalyst.fourmilab.ch/uploads/default/original/3X/f/a/fa93b2a5d9110c306cb7954093d618e89db2af02.png) ---
**Author:** ![CTLaw](https://scanalyst.fourmilab.ch/user_avatar/scanalyst.fourmilab.ch/ctlaw/32/42_2.png) [@CTLaw](https://scanalyst.fourmilab.ch/u/CTLaw)\ **Post date:** [19 October 2026 15:08 UTC](https://scanalyst.fourmilab.ch/t/generative-artificial-intelligence-large-language-models-and-image-synthesis/2383/880 "2026-10-19T15:08:18Z")
> > The defendant is a Saudi citizen who had been pursuing opportunities in Saudi Arabia before he was arrested. He had intended to fly from San Francisco to Riyadh (via Istanbul) with his wife and son on a one-way ticket on October 12, 2026. Having been arrested by the FBI three days prior, the defendant did not make that flight. His wife and son, however, immediately left the United States without him. Rather than wait until the October 12 flight as booked, they initially planned to fly to Jeddah, Saudi Arabia, on October 11 on a one-way ticket booked and paid for by the Saudi government. That flight was canceled, however, when a member of the Saudi military booked and paid for yet another one-way ticket for the defendant’s wife and son, this one business class to Doha, Qatar, on October 10, 2026. They did make that flight, departing San Francisco about 36 hours after the defendant was arrested. The day before she left the United States, the defendant’s wife told the FBI agents who interviewed her that her U.S. visa was expiring and that she and her son were _permanently_ moving back to Saudi Arabia—and that her husband, the defendant, eventually intended to do the same. > **[gov.uscourts.cand.480829.12.0\_1.pdf](https://storage.courtlistener.com/recap/gov.uscourts.cand.480829/gov.uscourts.cand.480829.12.0_1.pdf)** > > 253.71 KB ---
**Author:** ![Citizen\_bitcoin](https://scanalyst.fourmilab.ch/user_avatar/scanalyst.fourmilab.ch/citizen_bitcoin/32/19488_2.png) [@Citizen\_bitcoin](https://scanalyst.fourmilab.ch/u/Citizen_bitcoin)\ **Post date:** [20 October 2026 06:20 UTC](https://scanalyst.fourmilab.ch/t/generative-artificial-intelligence-large-language-models-and-image-synthesis/2383/881 "2026-10-20T06:20:51Z")
> _The defendant, Abdulrahman Alfozan, a now-former engineer at OpenAI, is charged with concealing a sophisticated recording device in an executive OpenAI conference room in an effort to secretly and unlawfully record the oral communications of the persons meeting in that room, in violation of 18 U.S.C. § 2511(1)(b). The government submits this memorandum in support of its motion that the defendant, a Saudi national with considerable resources and ties to Saudi Arabia, be detained pending trial. As alleged in detail in the Complaint (Dkt. 1), and as set forth herein, the defendant was mere days away from leaving the United States for Saudi Arabia when he was arrested._ ---
**Author:** ![CTLaw](https://scanalyst.fourmilab.ch/user_avatar/scanalyst.fourmilab.ch/ctlaw/32/42_2.png) [@CTLaw](https://scanalyst.fourmilab.ch/u/CTLaw)\ **Post date:** [2 November 2026 13:13 UTC](https://scanalyst.fourmilab.ch/t/generative-artificial-intelligence-large-language-models-and-image-synthesis/2383/882 "2026-11-02T13:13:35Z")
Use in 3d printing: [![](https://img.youtube.com/vi/TUqScwjnO_k/maxresdefault.jpg "I'm an Engineer. This ChatGPT Test Scared Me.") ](https://www.youtube.com/watch?v=TUqScwjnO_k) [Previous page](https://scanalyst.fourmilab.ch/t/generative-artificial-intelligence-large-language-models-and-image-synthesis/2383.md?page=43) [Next page](https://scanalyst.fourmilab.ch/t/generative-artificial-intelligence-large-language-models-and-image-synthesis/2383.md?page=45)