It is easier to explain the retail trade in Penny shares as a hobby rather than an investment. There is a game aspect, of course. Sometimes a double title, or decreases in half, or disappears when the CEO runs into Greece. This kind of event is however occasional, while amateur aspects are constant. Most of the pleasure given by Penny Stock seems to come from looking at literature, fighting on the forums and harassing interested parties.
Last year, FT Alphaville wrote a long article on Quantum Blockchain Technologies, a former dotcom incubator, hotel and restaurateur quoted by AIM, who claims to have discovered methods to optimize the mining of Bitcoin. The message has aroused more readers’ comments than any other article from the author in recent times. And this despite the fact that QBT has a market capitalization to the publication of just over 10 million pounds sterling.
Today, QBT’s market capitalization amounts to just over 20 million pounds sterling, the actions that jumped almost 200 percent following the announcement on Friday of a “revolutionary realization for its tool of Extraction of predictive models of artificial intelligence (“ia”) bitcoin “:
This tool, known as a method of Ai Oracle (“Ai Oracle”), now performs live from Bitcoin mining of current blockchain blockchain blocks (that’s to say., approximately 879,000 blocks, after the C method was “recycled” to manage the current blockchain blocks).
The company’s board of directors considers that this is an important step, since this exclusive technology was used in tests to exploit Bitcoin with a competitive advantage compared to the same material without the implementation of ‘Ia oracle. The competitive advantage material in the mining sector allowed by the Oracle IA can be obtained either by reducing the energy cost of mining by around 30 %; or (ii), accelerate the extraction speed with energy consumption and current costs with a hash rate of approximately 30 %.
We asked for a demonstration on Friday morning. QBT said we could go to Italy or, after a boost, offered his technical team to send us a demonstration video. We chose the second option. The video has still not arrived. If this is the case in the near future, we will update the message.
To summarize, Bitcoin uses a calculation lottery to determine which minor creates the next block for the blockchain and wins the reward. QBT says it can use AI and so on to estimate the probability of producing the winning block before the calculation is finished.
It is a controversial affirmation. The discovery of models in bitcoin mining could be important because it could reveal a bug in a cryptographic protocol used in many safety infrastructures. It would be less important if the models were a side effect of a network quantity rather than the protocol, but its inventors would always have a printing machine that they would do well to keep secret. (For the moment, the QBT money printing machine is AIM.)
QBT had previously declared that during laboratory tests, he was able to abandon blocks without any chance of saving “almost 50 % of the time”. Friday’s announcement is about improving the performance of 30 percent during live tests, although there are many qualifications.
Since the company deploys Ai Oracle on limited material resources which it currently has (that’s to say., an FPGA chip), the bitcoin shades extracted in pool are relatively minimal, but the same FPGA chip with the Oracle AI leads to an advantage of around 30 % compared to an FPGA chip without the AI Oracle version.
In other words, if the Oracle IA had been implemented on the Asic chips, the company estimates that this would increase the mining performance of an ASIC chip by around 30 %. Currently, QBT uses an FPGA, which has the hash power of a very small fraction of an ASIC, but which still has an improvement of around 30 %, which constitutes the key objective of the company , mainly for validation and demonstration purposes.
The QBT press release mentions that live tests began around the 879,000 block, suggesting that live channel experience lasted less than a week. (The chain was at around 880,000 blocks at the time of publication, therefore on cycles of 10 minutes, this represents around 1,000 blocks, each block producing 3.125 new bitcoins.)
Asic fleas are fast and effective but costly to manufacture, because they have only one objective, while FPGA chips swap efficiency against flexibility. The difficulty of the Bitcoin network reaches a record level, so FPGA tends to be used only to jump on vogue Shitcoin.
Even using the most recent ASIC Bitcoin, an isolated operator still has a chance on a quadrillion to exploit the following block. For this reason, professional minors smoothing the yields by pooling resources, which means that everyone receives a share of the rewards proportional to the power of treatment they bring to the pool. QBT’s live test uses one of these pools, which makes it more difficult to understand what is exactly tested.
One way of thinking about it would be to think of a lottery union in Jackpot only, in which a member of the union thinks can detect unsuccessful tickets after purchase. The lottery operator reimburses spoiled tickets before the print at 30 pence per book, the premonition is therefore a useful talent. The union member can request a partial refund for unreliable tickets, confirming after the draw he would have lost. Taking reimbursements before the draw would reduce the cost of loss.
Of course, there is an incalculable chance that any The ticket turns out to be the winner. Maybe only the union member think They have supernatural predictive capacities, in which case they eliminate random tickets.
If these spoiled and partially reimbursed tickets were excluded from the union total, this would mean a reduction in earnings. The benefit of 30 percent from the reimbursements would actually be equivalent to a loss of 70 percent – even if, given the small chances of winning the jackpot, such ineffectiveness could take a long time to do the average. And if the union did not know that the tickets were spoiled, the member could claim a share of the gains on the basis of the tickets he did not have. Anyway, it’s not good.
To be clear, we have no evidence suggesting that QBT tests are wrong. We have not seen any evidence anyway. Asked for details and examples of methodology, the CEO, executive president and QBT consultant, Francesco Gardin, replied by email:
The AI Oracle implementation tests on FPGA are only the tip of the iceberg. The specific Oracle AI was formed for several weeks to manage the current blockchain blocks, while the current version of the Most C model was developed for more than a year. The FPGA implementation of the specific Oracle that deals in real time the current blockchain Bitcoin blockchain is only living proof of a software version which, if it is not on the hardware, cannot Operating quickly enough to cope with the 10-minute delay to produce winning atmosphere above a minimum objective accepted by the pool. We exploit Satoshi, certainly not Bitcoins, with a single FPGA. What really matters are the improved performance of an FPGA, with the Oracle AI, with the same double Sha-256 ways, compared to the same FPGA without Oracle.
As for whether the results of a single knot in a mining pool of unknown size executed over an indefinite period could be considered statistically significant, Gardin said:
A pool, with a few exceptions, can only accept earnings superior to a minimum objective, otherwise it would be too demanding in terms of calculation and bandwidth. With the current FPGA with the AI Oracle implementation, we only have a limited number of duplicates SHA-256, taking into account the available silicon surface, but always sufficient to be “perceived” by the pool. Performance of 30 % is the measurement of double-256 double calculations avoided, hence an additional hash time available. There is a deterministic statistical proportion between the axing rate, the target and the winned winners.
(A spokesperson for the company added that QBT could be reluctant to speak to FTAV because our last article “quoted random third parties, who obviously had no knowledge of what society was trying to achieve and, consequently , have lowered the course of action by 50 %.
QBT also declared by RNS last week that it had made a patent application, “implementation of binary decision -making trees”. Nothing of this name is still in the databases of British or European patents. Citing the company’s patent lawyer, a spokesperson for QBT said: “The title, the applicant and the deposit date will be available in the coming weeks. No other details will be available before 18 months after the deposit. (That’s normal.)
The effectiveness of resources has not yet been a great concern in the generative AI, and the construction of models almost always occurs in the cloud rather than on the device. It will therefore be interesting to learn how all these inferences integrate on a simple chip whose power the requirement determines its value.
QBT invents its own jargon-as “double way Sha-256”, an expression that had not appeared on the internet before its RNS patent-so understanding its meaning can sometimes be delicate. What the company seems to say is that its Oracle AI adds between 1 and 4 % to the basic architecture of a chip as well as its energy needs. Making such light circuit would be, from a technical, incredible point of view. While waiting for proof, it may seem incredible.
Gardin said last week that the company “delivered its first irrefutable result”:
We believe that this could constitute a major advance for the entire Bitcoin mining industry, Car QBT has developed an Oracle IA which can either reduce the energy cost of mining, or increase the speed of mining mining at current energy costs, about 30 %. , what it can prove by live demonstrations.
Someone must pay for all this innovation, so that a concept of action which burns liquidity and which depends on the placement of actions for its survival may not want to emphasize the distinction between “can Prove “and” proved “. No doubt, however, that its amateur shareholders continue to have them for their money.
Additional readings/looks:
– The breakthrough of Quantum Blockchain in real -time mining (proactive webcast Investors)
-An Anglo-Italian company claims to have bitcoin. People have questions (FTAV)
