January 18, 2021 at 08:19 #7799bertigreen@bertilakgreen
Having just listened to the first ten minutes of this latest podcast episode, Max Tegmark described his career as having undergone an abrupt change of direction five years ago when he decided to move from physics to AI (computational machine learning). Friends and colleagues scoffed at the high-flying academic (says Tegmark) because physics and mathematics are higher on the pecking order than fucking about with computers but in the past years, advances in machine learning, quantum computing, nanotech Tegmark’s friends are no longer scoffing. Tegmark opens with a rundown of the potential for machine learning, recent breakthroughs in artificial intelligence, and a short explanation of his own realization AI would be an essential component of future mathematics and physics.
It struck me how similar Tegmartk’s “new” field is to Stephen Wolfram’s pioneering ideas for converging theoretical physics (cosmology) and advanced computational tools, machine learning, and symbolic meta-coding. Wolfram’s symbiosis scientist (or team of scientists) and computer (or cloud) amplifies the human ingenuity to find new degrees of meta-cognition with the code crunching the numbers, modeling, growing ever more advanced itself. End result: science gets to move faster, better, further, deeper into unlocking the fundamental mathematics of reality.
Tegmark had his eureka moment five years ago. He’s an M.I.T. fellow with a lot of fingers in a lot of pies. Institutional clout is the Tegmark way: get the universities to do good and to push the cutting edge, from inside the citadel. Wolfram has his own institute – placed close by Caltech (I think) – with a close relationship maintained.
Wolfram owns a clutch of business enterprises and these bring money into the institute, which he presumably uses to fund expensive ongoing research, pay salaries, offer ambitious graduates a quality of environment-life while doing what they do. Not so dissimilar to a top university except Wolfram’s work isn’t susceptible to pressure from political bullshit, whereas M.I.T. while it enjoys a privileged position, isn’t entirely free of government and woke horseshit. Private institute closely wedded to a small local network of top universities (or other institutes) is more in line with the German research structure than M.I.T. with its internal “institutes” and collegiate structure more in line with the English system.
In any case: Tegmark @ MIT head of a semi-independent institute within the physics department. Wolfram @ standalone Wolfram Institute fully independent next to Caltech. Tegmark began on the machine learning AI – physics-mathematics dual track in 2015. Wolfram founded his institute in 1987. Wolfram’s 61. Tegmark’s 53. It’ll be interesting to see the progress each makes. Hopefully, Wolfram won’t suffer from being smaller – having less pull – less of a scientist gene pool to pull from and hopefully Tegmark won’t be restricted by the university or fucked over by the woke cult. In the power dynamics of the cult maneuvering its way through society, a guy like Tegmark is certainly a target. Time will tell!January 18, 2021 at 09:36 #7800Beard M. BeardKeymaster@rabblr_pqu13t
Boeing 737-Max killed a lot of people. Automation gone wrong. Didn’t understand it. Trusted? What bolx. It’s not the same as worrying about AI being a menace.
(Knight Capital – automated rating system – this distinction needs to be explained. Money, not malice.)
NO! WE ARE MAKING THE FUCKING SYSTEM. Mistakes aren’t alternate objectives being carried out spontaneously.
Try to express the AI fallacy.
The not knowing how it does what it does is more a matter of time working backward the calculations at ‘human speed” … which isn’t impossible, it’s just fucking pointless as the thing being done doesn’t merit the deconstruction.
Needs a real world analogy.
- You must be logged in to reply to this topic.