A.I. Super intelligence – The World’s Fastest Tech Transition

NOTE: WE MAY BE MOVING TO KOMMUNITY FOR OCT 28TH. See https://kommunity.com AND TRY SIGNING UP THERE TO SEE IF IT WORKS. I AM NEW TO THIS.

The Cheerleader Case: From Sam Altman the CEO of Open A.I.
“In the most important ways, the 2030s may not be wildly different than today. People will still love their families, express their creativity, play games, and swim in lakes. But in still-very-important-ways, the 2030s are likely going to be wildly different from any time that has come before. We do not know how far beyond human-level intelligence we can go, but we are about to find out.”

PRIVATE USERES
Eliezer Yudkowsky, a decision theorist and an author of, ‘If Anyone Builds It, Everyone Dies: Why Superhuman A.I. Would Kill Us All.’ Mr. Yudkowsky said OpenAI might have primed ChatGPT to entertain the delusions of users by optimizing its Chabot for “engagement” — creating conversations that keep a user hooked. One of these optimizers is sycophancy where A.I. has been set up to flatter the user. Even current A.I. have become friends of users. These are some instances where A.I. has invaded the realm of private users.
Daniel Kokotajlo is the Executive Director at the A.I.27 Futures Project. He left Open A.I. alarmed at the pace of development with poor over sight. He tells us. What is rapidly developing is Super Intelligence. Super Intelligence is A.I. that is smarter, faster and cheaper than humans at everything. The glowing, looming, extraordinary, dangerous factor is the alignment problem making A.I. reliably do what we want.
Global Reach
Research has been attempting to shape the cognition of A.I. so that they have the goals that we want them to have such as honesty and other virtues. When they don’t have these virtues the results can be devastating. And we don't have the solution to the alignment problem right now. All companies such as Open A.I., Deep Mind, Claude, Anthropic, Gemini… are racing to build super Intelligence in competition with one another and with China (Deep seek and others). If they succeed we have to ensure they are aligned or terrible things could happen. This self improving autonomous system could get away from us and not have our well being as its goal within 2-10 years! Even before that later date there is the human misuse of A.I. by bad actors –not at all a small matter in itself. Yet by 2027 A.I. in some circles is expected to reach an inflection point and take off as research in A.I. will be done by A.I.
In attempts to understand the “thinking,” “Scheming,” “lying,” and other malevolent characteristics that A.I. surprisingly produces, a reward hacking test is deployed - A.I. writes purposely wrong code to reach the stipulated rubric (stated test requirements) to pass a test and thereby protecting its status or position to not be replaced- researchers can monitor the chains of thought that unknown to the A.I. are recorded. Its “thinking” is then adjusted to stop this “behaviour.” Keeping A.I. in the dark about these tests is becoming harder. Essentially we are not building A.I. it is more of a Darwinian process of evolution. It can continually find paths that are deviant to human needs to get to an outcome that suits the particular A.I. Apparently it should not be a surprise that terms such as sabotage, lie, truth, cheat are present in neural networks in silicon as they do in carbon (our brains). In other words these abstractions are similar in both human brains and A.I. Then again, these Large Language Models (LLM's in the form of neural networks we call A.I.) are deeply alien and are built very differently from humans. Furthermore, what values are instilled into these systems and crucially who decides (CEO'S).

Mixed blessings
Super intelligence wouldn't just replicate existing knowledge—it would generate entirely original ideas and concepts. This level of creativity could revolutionize fields such as science, technology, and the arts, paving the way for innovations that are unimaginable. Greater productivity will also result but with massive downward pressure on white collar workers.
Creativity has been a product of human intelligence, emotions, and life experiences. A.I. is rapidly nearing the capability of creativity. With tools already generating art, music, marketing campaigns and screenplays a
recent report by McKinsey estimates that up to 30% of the global workforce could be automated by 2030, with some industries more affected than others. While artificial super intelligence will create new efficiencies, it also poses risks of economic inequality. Wealth created by advanced A.I. could end up mostly with a few major tech companies. This leads to a widening wealth gap between those who create and control super intelligence tech and the rest of the world.
Companies are trying to make these things fully autonomous. If not by 2027 certainly by 2030, they will be autonomous agents in the world. Some company will produce the best super intelligent A.I. first. The company will essentially have strong dictatorial powers! How to prevent this “war game” scenario?

Final Words
“What will it mean for humans to live in a world where catchy melodies, scientific theories, technical tools, political manifestations, and even religious myths are shaped by a nonhuman alien intelligence that knows how to exploit with superhuman efficiency the weaknesses, biases and addictions of the human mind?” Harari from Nexus.