# Dear AI haters — very close, you’re only 180 degrees off Date: 2023-05-08 --- --- ### Dear AI haters — very close, you’re only 180 degrees off The logic of AI hating seems implacable: AI is a genie in a bottle and our only safe bet is keeping it there. The first problem with this logic is that humans are very bad at keeping things bottled — nuclear weapons, greenhouse gases and [gain of function viruses](https://www.marshall.senate.gov/wp-content/uploads/MWG-FDR-Document-04-11-23-EMBARGOED.pdf?fbclid=IwAR03xOoRDx1dYzpAkqxOvnw17ehBjRIfxSH9NQROn9KsF3KWx9Ss1us_ZAY) have already been released. The second problem with AI specific hating is that artificial intelligence isn’t even the biggest threat within intelligence research much less research in general. As [Captain Malcom Reynolds of Serenity](https://www.imdb.com/title/tt0379786/quotes/qt0433137) said, “A year from now, ten? They’ll swing back to the belief that they can make people… better.” Tinkering with humans for transhumanism purposes will always be the greater threat because messing with existing millions of years old technology is easier than building your own. So seen in this larger context the AI genie in a bottle logic falls apart because there are many bottles and some of the genies are much scarier. **AI isn’t the real threat; AI is our only hope of keeping technology from destroying humanity.** From the beginning of first thinking of AI in *WarGames* to *I, Robot* to *Skynet* having AI try to protect humanity has been portrayed as a disaster. But that’s because AI is almost always portrayed as a human like intelligence that thinks really fast. AI is not human like intelligence. Not only is human like intelligence currently impossible but the arrogance of thinking that any intelligence created will be similar to human intelligence is just silly. Silly because, not only are AI haters opposing diversity, they are refusing to admit that true diversity of intelligence could even exist. Saving the world is what current AI technology can do and it’s right in its wheelhouse. Provided with a good enough labeled dataset AI can judge the risk associated with all of the world’s research proposals. The system, let’s call it ResearchNet, would be open source and backed up and feed backed by human judgement for the close calls. Enforcement of the AI’s decisions would be handled the same way nuclear weapon research is — cross the line and it’s sanctions and exclusion from the World Trade Organization, etc. Okay now fine for the AI haters out there how come ResearchNet, after reading enough research proposals, doesn’t become sentient, gain human intelligence and therefore be a threat. Well for readers of my blog, [here](https://medium.com/@disrael/are-humans-sentient-5060a9c2824), [here](https://medium.com/@disrael/the-search-for-sentient-life-on-earth-47e70584495a) and [here](https://medium.com/@disrael/the-search-for-sentient-life-on-earth-rules-and-regulations-57901b78dfaa) the answer is simple: sentience is going beyond original programming and there are not even a lot of indications of humans accomplishing that much less our much less sophisticated AI. But say the AI haters what if you had a really sophisticated ResearchNet? First off ResearchNet would likely flag research producing that level of sophistication as illegal. Secondly, again AI intelligence is not human intelligence — the AI doesn’t have greed or self preservation built into it’s original program like a human. Let’s take an in depth look at the [open letter](https://futureoflife.org/open-letter/pause-giant-ai-experiments/) signed by so far by 27565 people. > Should we let machines flood our information channels with propaganda and untruth? Should we automate away all the jobs, including the fulfilling ones? Should we develop nonhuman minds that might eventually outnumber, outsmart, obsolete and replace us? Should we risk loss of control of our civilization? So taking these one by one. *Should we let machines flood our information channels with propaganda and untruth?* This one is again weirdly 180 degrees off. Our information channels are already flooded with propaganda and untruth. AI led fact checking might be our only hope of fixing that. Here’s Katherine J. Wu [shilling](https://www.theatlantic.com/science/archive/2023/03/covid-pandemic-origins-missing-evidence-debate/673460/) for a natural cause for COVID with the thoroughly debunked raccoon dogs and that’s a month after the government report above that debunks it! Anyone that believes an AI will any time soon be able to flood information channels with propaganda and untruth better than a human should seek scientific help. Humans brains have been training for 300,000 years to persuade other humans and some scientists believe that might even be the reason human intelligence developed. AI cannot any time soon compete with humans on lying but training to detect lies and provide a plausibility score might be possible — let’s call this one LieNet. *Should we automate away all the jobs, including the fulfilling ones? Should we develop nonhuman minds that might eventually outnumber, outsmart, obsolete and replace us?* Doing a twofer on these questions cause they seem to be the same. Cover your ears, cause the dog whistle on the “including the fulfilling ones” is LOUD. Automation has thus far meant that people that build houses, cook food, take care of children, and thousands of other jobs in the “real” economy make much less than those in FIRE, financial, insurance, and real estate, and any number of other “virtual” economy jobs at risk of being automated by a more sophisticated AI. First you don’t need AI technology to automate away these “virtual” jobs. Regular rules based software programming might be a slow way to automate FIRE jobs but it will get there. Even fully autonomous driving has an existing automation substitute — they’re called trains. Second from the AI haters FAQ > Many of the world’s leading AI scientists also think more powerful systems — often called “Artificial General Intelligence” (AGI) — that are competitive with humanity’s best at almost *any* task, are achievable, and this is the stated goal of many commercial AI labs. This is a lie so bold that we are back to needing ResearchNet and LieNet to protect us. Even if “best at almost *any* task” were near term possible that would just be one AI beating one human. But humans don’t work in groups of one we work in groups of thousands. So now that *any* task has to include the ability to communicate with thousands of other AI where each one is different and might provide inspiration. Sure such a thing might be possible many 100's of years from now but by then humanity could already have succumbed to the risk from other technological threats. *Should we risk loss of control of our civilization?* Again no we should not but as explained above AI seems our only hope of preventing that.