Finding our place in the age of AI | Oliver Edholm | TEDxSSE
HTML-код
- Опубликовано: 25 авг 2024
- Artificial intelligence is becoming smarter and smarter, we have figured out the equivalent of E=MC^2 of AI and scaling laws of development sees no end. Where do we find our place in all of this, as employees, entrepreneurs, and individuals?
Oliver Edholm shares his awakening and purpose in life with the goal of solving global issues with human-like intelligence in a computer, and shares the dangers of it without regulation.
At the age of 15, Oliver Edholm dropped out of high school and booked a one-way ticket to Singapore to research AI - without his parents knowing. Two years later, he founded his own AI startup Depict, receiving over $20 million in investments. Having worked at Klarna, done research on Artificial Intelligence and been included in the Forbes 30 under 30 list to add to his CV, Edholm is one of the youngest top pioneers within AI today at only 21 years old. This talk was given at a TEDx event using the TED conference format but independently organized by a local community. Learn more at www.ted.com/tedx
Good talk, but I think we missed the main conclusion: What the performer thinks our place will be in the age of AGI?
Very good talk 🎉
he has told so many interesting stuff!
Good intro effort for students - but a few inaccuracies, young and passionate.
The scaling law is NOT the equivalent of E=MC^2. The energy requirements that you mentioned will be a serious bottleneck. The true equivalent of Einstein's equation will be the invention of a more efficient form of processing (that will probably involve breakthroughs in materials engineering) - efficiency that lets us do the same level of compute for at least 100x less energy. The human brain does amazing compute while using less than a watt. Computers don't need to reach that level of efficiency (cuz they are faster and more reliable than organic brains in other ways), but they clearly DO need to reach at least one or two orders of magnitude increased efficiency. You overall timeline could still be roughly correct, but the road to AGI will need to be based to a great extent of efficiency of processing - not just more training data, etc.
FYI - the 7 Trillion figure is a Global effort over the next 10 years among many companies. He failed to mentioned this. ALso, this was just an estimate. In reality, could be as much as 2 or 3 Trillion over 15 years in a global effort among different countries.
This scaling law❓ Expensive. There was a trillion-dollar uncertainty in the sector. Viewed from increased productivity, one cannot take this all the way to the bank. It has yet to be proven. There are other problems with making this linearized claim.
I'm trying to release free-to-the-world plausible hope that civilization can quit the second law of thermodynamics. The second law is behind modern refgeration needing electrical energy to compress the refrigerent to force it to release as waste the heat that it has removed from the refrigerator's service interior in the cooling part of the refrigerent's circulation. There is also waste heat from mechanical friction.
Refrigeration by the principle that energy is conserved should produce electricity instead of consuming it.
It makes more sense rthat refrigerators should yield electricity because energy is widely known to change form with no ultimate path of energy gain or loss being found. Therefore any form of fully recyclable energy can be cycled endlessly in any quantity.
Full heat recycling all electric very isolated underground communities would be highly survivable with self sufficient light banks, farms, thaw resistant frozen food storehouses, factories, dwellings, and self contained elevators and horizontal transports.
If a high majority thinks our civilization should geoengineer gigatons or
teratons of carbon dioxide out of our environment, instalations using devices that convert ambient heat into electricity can hypothetically be scaled up do it with a choice of comsequences including many beneficial ones.
Computers that consume electricity and yield heat would complement energy sensible refrigerators that absorb heat and yield electricity. Computing would be free.
A simple rectifier crystal can, iust short of a replicatable long term demonstration of a powerful prototype, almost certainly filter the random thermal motion of electrons or discrete positiive charged voids called holes so the electric current flowing in one direction predominates. At low system voltage a filtrate of one polarity predominates only a little but there is always usable electrical power derived from the source, which is Johnson Nyquest thermal electrical noise. This net electrical filtrate can be aggregated in a group of separate diodes in consistent alignment parallel creating widely scalable electrical power. The maximum energy is converted from ambient heat to productive electricity when the electrical load is matched to the array impeadence.
Matched impeadence output (watts) is k (Boltzman's constant ~1.38^-23), times T (tempeature Kelvin) times bandwidth (0 Hz to a natural limit ~2 THz @ 290 K) times intrinsic (50%) and available rectification efficiency times the number of diodes in the array.
For reference, there are a billion cells of 1000 square nanometer area each per square millimeter, 100 billion per square centimeter.
Order is imposed on the random thermal motion of electrons by the structual orderlyness of a diode array made of diodes made within a slab:
v v v v v v v v v v v v v v v v v
All the P type semiconductor anodes abut a metal conductive plane deposited on the top face of the slab with nonrectifying joins; all the N type semiconductor cathodes abut the bottom face. As the polarity filtered electrical energy is exported, the amount of thermal energy in the group of diodes decreases. This group cooling will draw heat in from the surrounding ambient heat at a rate depending on the filtering rate and thermal resistance between the group and ambient gas, liquid, or solid warmer than absolute zero. There is a lot of ambient heat on our planet, more in equatorial dry desert summer days and less in polar desert winter nights.
Focusing on explaining the electronic behavior of one composition of simple diode, a near flawless crystal of silicon is modified by implanting a small amount of phosphorus (N type)on one side from a ohmic contact end to a junction where the additive is suddenly and completely changed to boron (P type) with minimal disturbance of the crystal pattern. The crystal then continues to another ohmic contact.
A region of high electrical resistance forms at the junction in this type of diode when the phosphorous near the ĵunction donates electrons that are free to move elsewhere while leaving phosphorus ions held in the crystal while the boron donates a hole which is similalarly free to move. The two types of mobile charges mutually clear each other away near the junction leaving little electrical conductivity. An equlibrium width of this region is settled between the phosphorus, boron, electrons, and holes. Thermal noise is beyond steady state equlibrium. Thermal noise transients, where mobile electrons move from the phosphorus added side to the boron added side ride transient extra conductivity so the forward moving electrons are preferentally filtered into the external circuit. Electrons are units of electric current. They lose their thermal energy of motion and gain electromotive force, another name for voltage, as they transition between the junction and the array electrical tap. Inside the diode, heat is absorbed: outside the diode, an attached electrical circuit is energized.
Understanding diodes is one way to become convinced that Johnson Nyquest thermal electrical noise can be rectified and aggregated. Development teams will find other ways to accomplish this wide mission. Taxonomically there should be many ways ways to convert heat directly into electricity.
A practical device may use an array of Au needles in a SiO2 matrix abutting N type GaAs. these were made in the 1970s when registration technology was poor so it was easier to fabricate arrays and select one diode than just make one.
There are other plausible breeches of the second law of thermodynamics. I hope a lot of people will join in expanding the breech. Please share the results of progress or setbacks.
These devices would probably become segmented commodities sold with minimal margin over supply cost. They would be manufactured by advanced automation that does not need financial incentive. Applicable best practices would be adopted. Business details would be open public knowledge. Associated people should move as negotiated and freely and honestly talk. Commerce would be a unified conglomerate of planetary scale of diverse local cooperatives. There is no need of wealth extracting top commanders. We do not need often token philanthropy from the top if the wide majority of people can afford to be generous.
Aloha
Charles M Brown
Kilauea Kauai Hawaii 96754
7 trillions dollars... Not yet 😏
this doesn't quite hit
Hei bro Are you looking for a professional youtube thumbnails designer and video SEO expert?