
Top 10 Web3 Crypto AI Agents
Onuora Amobi ·

On Tuesday, November 17, the BNB Chain prediction market Predict.fun postponed its PREDICT token generation event and Season 1 airdrop claims for the second time, a day before claims were due to open. The team said pushing ahead now would not be right for the token or its community, and pointed to a new window of two to four weeks, with no date. We looked at this airdrop's 35%-or-wait menu when it first went live in late October. What is new this week is who gets offered an exit. According to TechFlow's report of the announcement, participants in Binance Wallet's Prime Sale who do not want to keep waiting can file a refund application until 23:59 UTC on November 22.
Nobody announced anything comparable for the people holding Predict Points. Their option, per the same report, is that the claim-method selection stays open until TGE. They can keep switching between 35% now and the full amount over 12 months. They cannot leave.
One group paid in BNB and gets its BNB back. The other paid in trading volume and gets more time.
The timeline explains the whole thing. The Prime Sale ran for two hours on the morning of October 29, from 08:00 to 10:00 UTC, capped at 6 BNB per user. Binance had set PREDICT's Alpha debut for 09:00 UTC the following day and spot trading for 10:00 UTC, with USDT, USDC and TRY pairs.
At 20:05 UTC that evening, about ten hours after the sale closed, Binance hot wallets started emptying across ten blockchains. TRM Labs counted $6.36 billion taken, and Elliptic reports that the FBI has since attributed the theft to North Korea's TraderTraitor group. That same day Binance suspended Alpha trading, along with Alpha instant orders and spot-wallet token purchases, citing "ongoing security measures." Bloomingbit noted that Binance gave no timeline for full restoration.
PREDICT's listing went with it. Predict.fun delayed its TGE the next day, explicitly in line with the postponement of Alpha and spot trading. On November 7, as it published Season 2 points for 60,213 users, the project said the token remained on hold because Binance Wallet Alpha was still offline. Two days later Binance Alpha resumed withdrawals on BSC, Ethereum, Base and Arbitrum, restoring other networks "in stages." As of the latest reports we could find, Alpha trading itself has not come back.
The usual worry about exchange-led launches is commercial: listing fees, allocations set aside for the exchange's own users, a sale structure the project doesn't fully control. Predict.fun shows a plainer risk. When the token's sale, its airdrop claims, its first trading and its price discovery all run through one company's product, that company's bad day becomes the project's launch date.
Nothing in Predict.fun's code broke. Its supply of one billion tokens, the 70 million Season 1 airdrop pool and the 24-month insider cliff are the same as they were a month ago. What disappeared was the venue. And because the plan put that venue in the critical path for every step, the project had no clean way to go live without it. Launching on its own DEX liquidity while Binance's Prime Sale buyers sat waiting for tokens inside a frozen product would have been its own mess.
Compare the launch that did happen on schedule this week. infiniFi airdropped its INFI token into recipients' wallets on November 16 in a frozen, non-transferable state, then unfroze everything in a single transaction at 15:30 UTC on November 17, with Kraken named as a listing venue and others to follow. The tokens were already on-chain. The exchange was a place to trade them, not the place they lived.
Here is the asymmetry. A Prime Sale buyer committed BNB for a few hours on October 29 and was promised tokens delivered on TGE day. When TGE slipped twice, that buyer was offered their money back. A Season 1 user spent months providing the volume and liquidity that made PREDICT worth selling in a Prime Sale at all, and their allocation sits in the same holding pattern with no exit attached.
There is a reason for the difference, and it is not sinister. Buyers handed over an asset. Refunding it is mechanically simple, and leaving paid customers' money stuck inside a product that just lost billions is a liability no exchange wants. Points holders gave up nothing that can be handed back. You cannot refund someone's trading history.
But the effect is the same either way. The people with the least exposure to the project are the ones with the cleanest exit. The people with the most exposure are told to wait, and meanwhile the market keeps pricing what they're owed. On Aspecta's pre-market, where PREDICT has traded since October 9, the token hit $15.37 at launch and sat near $10.04 by November 2, about a third lower, before a single airdropped token could move. Season 1 users can see that price. They cannot act on it.
The strongest argument for Predict.fun's choice is a good one. Launching while Binance's Alpha product was offline would have split liquidity, stranded Prime Sale buyers and almost certainly produced a messy first day with thin books and wild prints. Airdrop recipients choosing the 35% instant option would have been selling into exactly that. Waiting protects them from the worst version of price discovery, and the open selection window means nobody is locked into a decision made before the delay.
It is also fair to say the project didn't cause any of this. Predict.fun picked the largest exchange in the world as its launch partner, which is not a reckless choice. Few teams plan for that partner losing $6 billion in 20 minutes.
All of that is true, and the thesis still holds. Waiting is defensible. Waiting with no date, two to four weeks after already waiting nearly three, with the exit offered only to the side that paid, tells users where they rank. It also tells every team planning a launch something more useful: decide in advance which parts of a token's life have to live on-chain and which can live inside a partner's product.
The pieces that matter most to holders, meaning the allocation, the vesting schedule and the claim, can sit in contracts that no exchange outage touches. Insider locks and vesting schedules held in public smart contracts keep running whether or not any single venue is online. Sales and listings can then plug into those contracts instead of replacing them, which is roughly the model a launchpad that settles on-chain is built around.
Predict.fun now says two to four weeks. The Prime Sale refund window closes at midnight UTC on Sunday. After that, everyone still in line is waiting on the same thing, and it isn't Predict.fun.

Onuora Amobi ·

Onuora Amobi ·

Onuora Amobi ·

Onuora Amobi ·

Onuora Amobi ·

Onuora Amobi ·